#include "llama-impl.h"
#include <map>
+#include <set>
static const std::map<llm_arch, const char *> LLM_ARCH_NAMES = {
{ LLM_ARCH_CLIP, "clip" }, // dummy, only used by llama-quantize
{ LLM_ARCH_JAIS, "jais" },
{ LLM_ARCH_NEMOTRON, "nemotron" },
{ LLM_ARCH_NEMOTRON_H, "nemotron_h" },
+ { LLM_ARCH_NEMOTRON_H_MOE, "nemotron_h_moe" },
{ LLM_ARCH_EXAONE, "exaone" },
{ LLM_ARCH_EXAONE4, "exaone4" },
{ LLM_ARCH_RWKV6, "rwkv6" },
{ LLM_KV_TOKENIZER_MIDDLE_ID, "tokenizer.ggml.middle_token_id" },
};
-static const std::map<llm_arch, std::map<llm_tensor, const char *>> LLM_TENSOR_NAMES = {
- {
- LLM_ARCH_CLIP,
- {},
- },
- {
- LLM_ARCH_LLAMA,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ROPE_FREQS, "rope_freqs" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_ATTN_ROT_EMBD, "blk.%d.attn_rot_embd" },
- { LLM_TENSOR_FFN_GATE_INP, "blk.%d.ffn_gate_inp" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- { LLM_TENSOR_FFN_GATE_EXP, "blk.%d.ffn_gate.%d" },
- { LLM_TENSOR_FFN_DOWN_EXP, "blk.%d.ffn_down.%d" },
- { LLM_TENSOR_FFN_UP_EXP, "blk.%d.ffn_up.%d" },
- { LLM_TENSOR_FFN_GATE_EXPS, "blk.%d.ffn_gate_exps" },
- { LLM_TENSOR_FFN_DOWN_EXPS, "blk.%d.ffn_down_exps" },
- { LLM_TENSOR_FFN_UP_EXPS, "blk.%d.ffn_up_exps" },
- },
- },
- {
- LLM_ARCH_ARCEE,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ROPE_FREQS, "rope_freqs" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_ATTN_ROT_EMBD, "blk.%d.attn_rot_embd" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- },
- },
- {
- LLM_ARCH_AFMOE,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_POST_NORM, "blk.%d.post_attention_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_ATTN_Q_NORM, "blk.%d.attn_q_norm" },
- { LLM_TENSOR_ATTN_K_NORM, "blk.%d.attn_k_norm" },
- { LLM_TENSOR_ATTN_GATE, "blk.%d.attn_gate" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_POST_NORM, "blk.%d.post_ffw_norm" },
- { LLM_TENSOR_FFN_GATE_INP, "blk.%d.ffn_gate_inp" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- { LLM_TENSOR_FFN_GATE_EXPS, "blk.%d.ffn_gate_exps" },
- { LLM_TENSOR_FFN_DOWN_EXPS, "blk.%d.ffn_down_exps" },
- { LLM_TENSOR_FFN_UP_EXPS, "blk.%d.ffn_up_exps" },
- { LLM_TENSOR_FFN_GATE_SHEXP, "blk.%d.ffn_gate_shexp" },
- { LLM_TENSOR_FFN_UP_SHEXP, "blk.%d.ffn_up_shexp" },
- { LLM_TENSOR_FFN_DOWN_SHEXP, "blk.%d.ffn_down_shexp" },
- { LLM_TENSOR_FFN_EXP_PROBS_B, "blk.%d.exp_probs_b" },
- },
- },
- {
- LLM_ARCH_LLAMA4,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ROPE_FREQS, "rope_freqs" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_ATTN_ROT_EMBD, "blk.%d.attn_rot_embd" },
- { LLM_TENSOR_FFN_GATE_INP, "blk.%d.ffn_gate_inp" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- { LLM_TENSOR_FFN_GATE_EXP, "blk.%d.ffn_gate.%d" },
- { LLM_TENSOR_FFN_DOWN_EXP, "blk.%d.ffn_down.%d" },
- { LLM_TENSOR_FFN_UP_EXP, "blk.%d.ffn_up.%d" },
- { LLM_TENSOR_FFN_GATE_EXPS, "blk.%d.ffn_gate_exps" },
- { LLM_TENSOR_FFN_DOWN_EXPS, "blk.%d.ffn_down_exps" },
- { LLM_TENSOR_FFN_UP_EXPS, "blk.%d.ffn_up_exps" },
- { LLM_TENSOR_FFN_GATE_SHEXP, "blk.%d.ffn_gate_shexp" },
- { LLM_TENSOR_FFN_DOWN_SHEXP, "blk.%d.ffn_down_shexp" },
- { LLM_TENSOR_FFN_UP_SHEXP, "blk.%d.ffn_up_shexp" },
- },
- },
- {
- LLM_ARCH_DECI,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ROPE_FREQS, "rope_freqs" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_ATTN_ROT_EMBD, "blk.%d.attn_rot_embd" },
- { LLM_TENSOR_FFN_GATE_INP, "blk.%d.ffn_gate_inp" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- { LLM_TENSOR_FFN_GATE_EXP, "blk.%d.ffn_gate.%d" },
- { LLM_TENSOR_FFN_DOWN_EXP, "blk.%d.ffn_down.%d" },
- { LLM_TENSOR_FFN_UP_EXP, "blk.%d.ffn_up.%d" },
- { LLM_TENSOR_FFN_GATE_EXPS, "blk.%d.ffn_gate_exps" },
- { LLM_TENSOR_FFN_DOWN_EXPS, "blk.%d.ffn_down_exps" },
- { LLM_TENSOR_FFN_UP_EXPS, "blk.%d.ffn_up_exps" },
- },
- },
- {
- LLM_ARCH_BAICHUAN,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ROPE_FREQS, "rope_freqs" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_ATTN_ROT_EMBD, "blk.%d.attn_rot_embd" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- },
- },
- {
- LLM_ARCH_FALCON,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_NORM_2, "blk.%d.attn_norm_2" },
- { LLM_TENSOR_ATTN_QKV, "blk.%d.attn_qkv" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- },
- },
- {
- LLM_ARCH_GROK,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ROPE_FREQS, "rope_freqs" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_ATTN_ROT_EMBD, "blk.%d.attn_rot_embd" },
- { LLM_TENSOR_FFN_GATE_INP, "blk.%d.ffn_gate_inp" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- { LLM_TENSOR_FFN_GATE_EXP, "blk.%d.ffn_gate.%d" },
- { LLM_TENSOR_FFN_DOWN_EXP, "blk.%d.ffn_down.%d" },
- { LLM_TENSOR_FFN_UP_EXP, "blk.%d.ffn_up.%d" },
- { LLM_TENSOR_FFN_GATE_EXPS, "blk.%d.ffn_gate_exps" },
- { LLM_TENSOR_FFN_DOWN_EXPS, "blk.%d.ffn_down_exps" },
- { LLM_TENSOR_FFN_UP_EXPS, "blk.%d.ffn_up_exps" },
- { LLM_TENSOR_FFN_POST_NORM, "blk.%d.post_ffw_norm" },
- { LLM_TENSOR_LAYER_OUT_NORM, "blk.%d.layer_output_norm" },
- { LLM_TENSOR_ATTN_OUT_NORM, "blk.%d.attn_output_norm" },
- },
- },
- {
- LLM_ARCH_GPT2,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_POS_EMBD, "position_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_QKV, "blk.%d.attn_qkv" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- },
- },
- {
- LLM_ARCH_GPTJ,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- },
- },
- {
- LLM_ARCH_GPTNEOX,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_QKV, "blk.%d.attn_qkv" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- },
- },
- {
- LLM_ARCH_MPT,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output"},
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_ATTN_QKV, "blk.%d.attn_qkv" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- { LLM_TENSOR_FFN_ACT, "blk.%d.ffn.act" },
- { LLM_TENSOR_POS_EMBD, "position_embd" },
- { LLM_TENSOR_ATTN_Q_NORM, "blk.%d.attn_q_norm"},
- { LLM_TENSOR_ATTN_K_NORM, "blk.%d.attn_k_norm"},
- },
- },
- {
- LLM_ARCH_STARCODER,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_POS_EMBD, "position_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_QKV, "blk.%d.attn_qkv" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- },
- },
- {
- LLM_ARCH_REFACT,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- },
- },
- {
- LLM_ARCH_BERT,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_TOKEN_EMBD_NORM, "token_embd_norm" },
- { LLM_TENSOR_TOKEN_TYPES, "token_types" },
- { LLM_TENSOR_POS_EMBD, "position_embd" },
- { LLM_TENSOR_ATTN_OUT_NORM, "blk.%d.attn_output_norm" },
- { LLM_TENSOR_ATTN_QKV, "blk.%d.attn_qkv" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_LAYER_OUT_NORM, "blk.%d.layer_output_norm" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- { LLM_TENSOR_CLS, "cls" },
- { LLM_TENSOR_CLS_OUT, "cls.output" },
- },
- },
- {
- LLM_ARCH_NOMIC_BERT,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_TOKEN_EMBD_NORM, "token_embd_norm" },
- { LLM_TENSOR_TOKEN_TYPES, "token_types" },
- { LLM_TENSOR_ATTN_OUT_NORM, "blk.%d.attn_output_norm" },
- { LLM_TENSOR_ATTN_QKV, "blk.%d.attn_qkv" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_LAYER_OUT_NORM, "blk.%d.layer_output_norm" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- },
- },
- {
- LLM_ARCH_NOMIC_BERT_MOE,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_TOKEN_EMBD_NORM, "token_embd_norm" },
- { LLM_TENSOR_TOKEN_TYPES, "token_types" },
- { LLM_TENSOR_ATTN_OUT_NORM, "blk.%d.attn_output_norm" },
- { LLM_TENSOR_ATTN_QKV, "blk.%d.attn_qkv" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_LAYER_OUT_NORM, "blk.%d.layer_output_norm" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- { LLM_TENSOR_FFN_GATE_INP, "blk.%d.ffn_gate_inp" },
- { LLM_TENSOR_FFN_DOWN_EXPS, "blk.%d.ffn_down_exps" },
- { LLM_TENSOR_FFN_UP_EXPS, "blk.%d.ffn_up_exps" },
- },
- },
- {
- LLM_ARCH_NEO_BERT,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_QKV, "blk.%d.attn_qkv" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- { LLM_TENSOR_ENC_OUTPUT_NORM, "enc.output_norm" },
- { LLM_TENSOR_CLS, "cls" },
- { LLM_TENSOR_CLS_OUT, "cls.output" },
- },
- },
- {
- LLM_ARCH_JINA_BERT_V2,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_TOKEN_EMBD_NORM, "token_embd_norm" },
- { LLM_TENSOR_TOKEN_TYPES, "token_types" },
- { LLM_TENSOR_ATTN_NORM_2, "blk.%d.attn_norm_2" },
- { LLM_TENSOR_ATTN_OUT_NORM, "blk.%d.attn_output_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_Q_NORM, "blk.%d.attn_q_norm" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_K_NORM, "blk.%d.attn_k_norm" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_LAYER_OUT_NORM, "blk.%d.layer_output_norm" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- { LLM_TENSOR_CLS, "cls" },
- },
- },
- {
- LLM_ARCH_JINA_BERT_V3,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_TOKEN_EMBD_NORM, "token_embd_norm" },
- { LLM_TENSOR_TOKEN_TYPES, "token_types" },
- { LLM_TENSOR_ATTN_OUT_NORM, "blk.%d.attn_output_norm" },
- { LLM_TENSOR_ATTN_QKV, "blk.%d.attn_qkv" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- { LLM_TENSOR_LAYER_OUT_NORM, "blk.%d.layer_output_norm" },
- },
- },
- {
- LLM_ARCH_BLOOM,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_TOKEN_EMBD_NORM, "token_embd_norm" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_QKV, "blk.%d.attn_qkv" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- },
- },
- {
- LLM_ARCH_STABLELM,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ROPE_FREQS, "rope_freqs" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- { LLM_TENSOR_ATTN_Q_NORM, "blk.%d.attn_q_norm" },
- { LLM_TENSOR_ATTN_K_NORM, "blk.%d.attn_k_norm" },
- },
- },
- {
- LLM_ARCH_QWEN,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ROPE_FREQS, "rope_freqs" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_QKV, "blk.%d.attn_qkv" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- },
- },
- {
- LLM_ARCH_QWEN2,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- },
- },
- {
- LLM_ARCH_QWEN2VL,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- },
- },
- {
- LLM_ARCH_QWEN2MOE,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_GATE_INP, "blk.%d.ffn_gate_inp" },
- { LLM_TENSOR_FFN_GATE_EXPS, "blk.%d.ffn_gate_exps" },
- { LLM_TENSOR_FFN_DOWN_EXPS, "blk.%d.ffn_down_exps" },
- { LLM_TENSOR_FFN_UP_EXPS, "blk.%d.ffn_up_exps" },
- { LLM_TENSOR_FFN_GATE_INP_SHEXP, "blk.%d.ffn_gate_inp_shexp" },
- { LLM_TENSOR_FFN_GATE_SHEXP, "blk.%d.ffn_gate_shexp" },
- { LLM_TENSOR_FFN_DOWN_SHEXP, "blk.%d.ffn_down_shexp" },
- { LLM_TENSOR_FFN_UP_SHEXP, "blk.%d.ffn_up_shexp" },
- },
- },
- {
- LLM_ARCH_QWEN3,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_CLS_OUT, "cls.output" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_Q_NORM, "blk.%d.attn_q_norm" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_K_NORM, "blk.%d.attn_k_norm" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- },
- },
- {
- LLM_ARCH_QWEN3MOE,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_Q_NORM, "blk.%d.attn_q_norm" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_K_NORM, "blk.%d.attn_k_norm" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_GATE_INP, "blk.%d.ffn_gate_inp" },
- { LLM_TENSOR_FFN_GATE_EXPS, "blk.%d.ffn_gate_exps" },
- { LLM_TENSOR_FFN_DOWN_EXPS, "blk.%d.ffn_down_exps" },
- { LLM_TENSOR_FFN_UP_EXPS, "blk.%d.ffn_up_exps" },
- },
- },
- {
- LLM_ARCH_QWEN3NEXT,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_POST_NORM, "blk.%d.post_attention_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_Q_NORM, "blk.%d.attn_q_norm" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_K_NORM, "blk.%d.attn_k_norm" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_GATE_INP, "blk.%d.ffn_gate_inp" },
- { LLM_TENSOR_FFN_GATE_EXPS, "blk.%d.ffn_gate_exps" },
- { LLM_TENSOR_FFN_DOWN_EXPS, "blk.%d.ffn_down_exps" },
- { LLM_TENSOR_FFN_UP_EXPS, "blk.%d.ffn_up_exps" },
- { LLM_TENSOR_FFN_GATE_INP_SHEXP, "blk.%d.ffn_gate_inp_shexp" },
- { LLM_TENSOR_FFN_GATE_SHEXP, "blk.%d.ffn_gate_shexp" },
- { LLM_TENSOR_FFN_DOWN_SHEXP, "blk.%d.ffn_down_shexp" },
- { LLM_TENSOR_FFN_UP_SHEXP, "blk.%d.ffn_up_shexp" },
- { LLM_TENSOR_SSM_A_NOSCAN, "blk.%d.ssm_a" },
- { LLM_TENSOR_SSM_CONV1D, "blk.%d.ssm_conv1d" },
- { LLM_TENSOR_SSM_DT, "blk.%d.ssm_dt" },
- { LLM_TENSOR_SSM_BETA_ALPHA, "blk.%d.ssm_ba" },
- { LLM_TENSOR_SSM_IN, "blk.%d.ssm_in" },
- { LLM_TENSOR_SSM_NORM, "blk.%d.ssm_norm" },
- { LLM_TENSOR_SSM_OUT, "blk.%d.ssm_out" },
- },
- },
- {
- LLM_ARCH_QWEN3VL,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_Q_NORM, "blk.%d.attn_q_norm" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_K_NORM, "blk.%d.attn_k_norm" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- },
- },
- {
- LLM_ARCH_QWEN3VLMOE,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_Q_NORM, "blk.%d.attn_q_norm" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_K_NORM, "blk.%d.attn_k_norm" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_GATE_INP, "blk.%d.ffn_gate_inp" },
- { LLM_TENSOR_FFN_GATE_EXPS, "blk.%d.ffn_gate_exps" },
- { LLM_TENSOR_FFN_DOWN_EXPS, "blk.%d.ffn_down_exps" },
- { LLM_TENSOR_FFN_UP_EXPS, "blk.%d.ffn_up_exps" },
- },
- },
- {
- LLM_ARCH_PHI2,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_QKV, "blk.%d.attn_qkv" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- },
- },
- {
- LLM_ARCH_PHI3,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ROPE_FACTORS_LONG, "rope_factors_long" },
- { LLM_TENSOR_ROPE_FACTORS_SHORT, "rope_factors_short" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_QKV, "blk.%d.attn_qkv" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- },
- },
- {
- LLM_ARCH_PHIMOE,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ROPE_FACTORS_LONG, "rope_factors_long" },
- { LLM_TENSOR_ROPE_FACTORS_SHORT, "rope_factors_short" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_QKV, "blk.%d.attn_qkv" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_GATE_INP, "blk.%d.ffn_gate_inp" },
- { LLM_TENSOR_FFN_GATE_EXPS, "blk.%d.ffn_gate_exps" },
- { LLM_TENSOR_FFN_DOWN_EXPS, "blk.%d.ffn_down_exps" },
- { LLM_TENSOR_FFN_UP_EXPS, "blk.%d.ffn_up_exps" },
- },
- },
- {
- LLM_ARCH_PLAMO,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ROPE_FREQS, "rope_freqs" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_ATTN_ROT_EMBD, "blk.%d.attn_rot_embd" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- },
- },
- {
- LLM_ARCH_PLAMO2,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ROPE_FREQS, "rope_freqs" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_QKV, "blk.%d.attn_qkv" },
- { LLM_TENSOR_ATTN_Q_NORM, "blk.%d.attn_q_norm" },
- { LLM_TENSOR_ATTN_K_NORM, "blk.%d.attn_k_norm" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_ATTN_ROT_EMBD, "blk.%d.attn_rot_embd" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- { LLM_TENSOR_SSM_IN, "blk.%d.ssm_in" },
- { LLM_TENSOR_SSM_CONV1D, "blk.%d.ssm_conv1d" },
- { LLM_TENSOR_SSM_X, "blk.%d.ssm_x" },
- { LLM_TENSOR_SSM_DT, "blk.%d.ssm_dt" },
- { LLM_TENSOR_SSM_A, "blk.%d.ssm_a" },
- { LLM_TENSOR_SSM_D, "blk.%d.ssm_d" },
- { LLM_TENSOR_SSM_OUT, "blk.%d.ssm_out" },
- { LLM_TENSOR_SSM_DT_NORM, "blk.%d.ssm_dt_norm" },
- { LLM_TENSOR_SSM_B_NORM, "blk.%d.ssm_b_norm" },
- { LLM_TENSOR_SSM_C_NORM, "blk.%d.ssm_c_norm" },
- { LLM_TENSOR_ATTN_POST_NORM, "blk.%d.post_attention_norm" },
- { LLM_TENSOR_FFN_POST_NORM, "blk.%d.post_ffw_norm" },
- },
- },
- {
- LLM_ARCH_CODESHELL,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ROPE_FREQS, "rope_freqs" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_QKV, "blk.%d.attn_qkv" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_ATTN_ROT_EMBD, "blk.%d.attn_rot_embd" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- },
- },
- {
- LLM_ARCH_ORION,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ROPE_FREQS, "rope_freqs" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_ATTN_ROT_EMBD, "blk.%d.attn_rot_embd" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- },
- },
- {
- LLM_ARCH_INTERNLM2,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- },
- },
- {
- LLM_ARCH_MINICPM,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ROPE_FREQS, "rope_freqs" },
- { LLM_TENSOR_ROPE_FACTORS_LONG, "rope_factors_long" },
- { LLM_TENSOR_ROPE_FACTORS_SHORT, "rope_factors_short" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_ATTN_ROT_EMBD, "blk.%d.attn_rot_embd" },
- { LLM_TENSOR_FFN_GATE_INP, "blk.%d.ffn_gate_inp" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- { LLM_TENSOR_FFN_GATE_EXP, "blk.%d.ffn_gate.%d" },
- { LLM_TENSOR_FFN_DOWN_EXP, "blk.%d.ffn_down.%d" },
- { LLM_TENSOR_FFN_UP_EXP, "blk.%d.ffn_up.%d" },
- },
- },
- {
- LLM_ARCH_MINICPM3,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ROPE_FACTORS_LONG, "rope_factors_long" },
- { LLM_TENSOR_ROPE_FACTORS_SHORT, "rope_factors_short" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q_A_NORM, "blk.%d.attn_q_a_norm" },
- { LLM_TENSOR_ATTN_KV_A_NORM, "blk.%d.attn_kv_a_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_Q_A, "blk.%d.attn_q_a" },
- { LLM_TENSOR_ATTN_Q_B, "blk.%d.attn_q_b" },
- { LLM_TENSOR_ATTN_KV_A_MQA, "blk.%d.attn_kv_a_mqa" },
- { LLM_TENSOR_ATTN_KV_B, "blk.%d.attn_kv_b" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- },
- },
- {
- LLM_ARCH_GEMMA,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- },
- },
- {
- LLM_ARCH_GEMMA2,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_ATTN_POST_NORM, "blk.%d.post_attention_norm" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- { LLM_TENSOR_FFN_POST_NORM, "blk.%d.post_ffw_norm" },
- },
- },
- {
- LLM_ARCH_GEMMA3,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_Q_NORM, "blk.%d.attn_q_norm" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_K_NORM, "blk.%d.attn_k_norm" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_ATTN_POST_NORM, "blk.%d.post_attention_norm" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- { LLM_TENSOR_FFN_POST_NORM, "blk.%d.post_ffw_norm" },
- },
- },
- {
- LLM_ARCH_GEMMA3N,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_Q_NORM, "blk.%d.attn_q_norm" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_K_NORM, "blk.%d.attn_k_norm" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_ATTN_POST_NORM, "blk.%d.post_attention_norm" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- { LLM_TENSOR_FFN_POST_NORM, "blk.%d.post_ffw_norm" },
- { LLM_TENSOR_PER_LAYER_TOKEN_EMBD, "per_layer_token_embd" },
- { LLM_TENSOR_PER_LAYER_MODEL_PROJ, "per_layer_model_proj" },
- { LLM_TENSOR_PER_LAYER_PROJ_NORM, "per_layer_proj_norm" },
- { LLM_TENSOR_ALTUP_UNEMBD_PROJ, "altup_unembd_proj" },
- { LLM_TENSOR_ALTUP_PROJ, "altup_proj" },
- { LLM_TENSOR_PER_LAYER_INP_GATE, "blk.%d.inp_gate" },
- { LLM_TENSOR_PER_LAYER_PROJ, "blk.%d.proj" },
- { LLM_TENSOR_PER_LAYER_POST_NORM, "blk.%d.post_norm" },
- { LLM_TENSOR_ALTUP_CORRECT_COEF, "blk.%d.altup_correct_coef" },
- { LLM_TENSOR_ALTUP_CORRECT_SCALE, "blk.%d.altup_correct_scale" },
- { LLM_TENSOR_ALTUP_PREDICT_COEF, "blk.%d.altup_predict_coef" },
- { LLM_TENSOR_ALTUP_ROUTER, "blk.%d.altup_router" },
- { LLM_TENSOR_ALTUP_ROUTER_NORM, "blk.%d.altup_router_norm" },
- { LLM_TENSOR_LAUREL_L, "blk.%d.laurel_l" },
- { LLM_TENSOR_LAUREL_R, "blk.%d.laurel_r" },
- { LLM_TENSOR_LAUREL_POST_NORM, "blk.%d.laurel_post_norm" },
- },
- },
- {
- LLM_ARCH_GEMMA_EMBEDDING,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_DENSE_2_OUT, "dense_2" },
- { LLM_TENSOR_DENSE_3_OUT, "dense_3" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_Q_NORM, "blk.%d.attn_q_norm" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_K_NORM, "blk.%d.attn_k_norm" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_ATTN_POST_NORM, "blk.%d.post_attention_norm" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- { LLM_TENSOR_FFN_POST_NORM, "blk.%d.post_ffw_norm" },
- },
- },
- {
- LLM_ARCH_STARCODER2,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ROPE_FREQS, "rope_freqs" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_ATTN_ROT_EMBD, "blk.%d.attn_rot_embd" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- },
- },
- {
- LLM_ARCH_MAMBA,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_SSM_IN, "blk.%d.ssm_in" },
- { LLM_TENSOR_SSM_CONV1D, "blk.%d.ssm_conv1d" },
- { LLM_TENSOR_SSM_X, "blk.%d.ssm_x" },
- { LLM_TENSOR_SSM_DT, "blk.%d.ssm_dt" },
- { LLM_TENSOR_SSM_A, "blk.%d.ssm_a" },
- { LLM_TENSOR_SSM_D, "blk.%d.ssm_d" },
- { LLM_TENSOR_SSM_OUT, "blk.%d.ssm_out" },
- },
- },
- {
- LLM_ARCH_MAMBA2,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_SSM_IN, "blk.%d.ssm_in" },
- { LLM_TENSOR_SSM_CONV1D, "blk.%d.ssm_conv1d" },
- { LLM_TENSOR_SSM_DT, "blk.%d.ssm_dt" },
- { LLM_TENSOR_SSM_A, "blk.%d.ssm_a" },
- { LLM_TENSOR_SSM_D, "blk.%d.ssm_d" },
- { LLM_TENSOR_SSM_NORM, "blk.%d.ssm_norm" },
- { LLM_TENSOR_SSM_OUT, "blk.%d.ssm_out" },
- },
- },
- {
- LLM_ARCH_JAMBA,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_SSM_IN, "blk.%d.ssm_in" },
- { LLM_TENSOR_SSM_CONV1D, "blk.%d.ssm_conv1d" },
- { LLM_TENSOR_SSM_X, "blk.%d.ssm_x" },
- { LLM_TENSOR_SSM_DT, "blk.%d.ssm_dt" },
- { LLM_TENSOR_SSM_DT_NORM, "blk.%d.ssm_dt_norm" },
- { LLM_TENSOR_SSM_A, "blk.%d.ssm_a" },
- { LLM_TENSOR_SSM_B_NORM, "blk.%d.ssm_b_norm" },
- { LLM_TENSOR_SSM_C_NORM, "blk.%d.ssm_c_norm" },
- { LLM_TENSOR_SSM_D, "blk.%d.ssm_d" },
- { LLM_TENSOR_SSM_OUT, "blk.%d.ssm_out" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_FFN_GATE_INP, "blk.%d.ffn_gate_inp" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- { LLM_TENSOR_FFN_GATE_EXPS, "blk.%d.ffn_gate_exps" },
- { LLM_TENSOR_FFN_DOWN_EXPS, "blk.%d.ffn_down_exps" },
- { LLM_TENSOR_FFN_UP_EXPS, "blk.%d.ffn_up_exps" },
- },
- },
- {
- LLM_ARCH_FALCON_H1,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_SSM_IN, "blk.%d.ssm_in" },
- { LLM_TENSOR_SSM_CONV1D, "blk.%d.ssm_conv1d" },
- { LLM_TENSOR_SSM_DT, "blk.%d.ssm_dt" },
- { LLM_TENSOR_SSM_A, "blk.%d.ssm_a" },
- { LLM_TENSOR_SSM_D, "blk.%d.ssm_d" },
- { LLM_TENSOR_SSM_NORM, "blk.%d.ssm_norm" },
- { LLM_TENSOR_SSM_OUT, "blk.%d.ssm_out" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- },
- },
- {
- LLM_ARCH_XVERSE,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ROPE_FREQS, "rope_freqs" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_ATTN_ROT_EMBD, "blk.%d.attn_rot_embd" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- },
- },
- {
- LLM_ARCH_COMMAND_R,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- { LLM_TENSOR_ATTN_Q_NORM, "blk.%d.attn_q_norm" },
- { LLM_TENSOR_ATTN_K_NORM, "blk.%d.attn_k_norm" },
- },
- },
- {
- LLM_ARCH_COHERE2,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- },
- },
- {
- LLM_ARCH_DBRX,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ATTN_QKV, "blk.%d.attn_qkv" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_ATTN_OUT_NORM, "blk.%d.attn_output_norm" },
- { LLM_TENSOR_FFN_GATE_INP, "blk.%d.ffn_gate_inp" },
- { LLM_TENSOR_FFN_GATE_EXPS, "blk.%d.ffn_gate_exps" },
- { LLM_TENSOR_FFN_DOWN_EXPS, "blk.%d.ffn_down_exps" },
- { LLM_TENSOR_FFN_UP_EXPS, "blk.%d.ffn_up_exps" },
- },
- },
- {
- LLM_ARCH_OLMO,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- },
- },
- {
- LLM_ARCH_OLMO2,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_ATTN_POST_NORM, "blk.%d.post_attention_norm" },
- { LLM_TENSOR_ATTN_Q_NORM, "blk.%d.attn_q_norm" },
- { LLM_TENSOR_ATTN_K_NORM, "blk.%d.attn_k_norm" },
- { LLM_TENSOR_FFN_POST_NORM, "blk.%d.post_ffw_norm" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- },
- },
- {
- LLM_ARCH_OLMOE,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_ATTN_Q_NORM, "blk.%d.attn_q_norm" },
- { LLM_TENSOR_ATTN_K_NORM, "blk.%d.attn_k_norm" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_GATE_INP, "blk.%d.ffn_gate_inp" },
- { LLM_TENSOR_FFN_GATE_EXPS, "blk.%d.ffn_gate_exps" },
- { LLM_TENSOR_FFN_DOWN_EXPS, "blk.%d.ffn_down_exps" },
- { LLM_TENSOR_FFN_UP_EXPS, "blk.%d.ffn_up_exps" },
- },
- },
- {
- LLM_ARCH_OPENELM,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_QKV, "blk.%d.attn_qkv" },
- { LLM_TENSOR_ATTN_Q_NORM, "blk.%d.attn_q_norm" },
- { LLM_TENSOR_ATTN_K_NORM, "blk.%d.attn_k_norm" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- },
- },
- {
- LLM_ARCH_ARCTIC,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_FFN_GATE_INP, "blk.%d.ffn_gate_inp" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- { LLM_TENSOR_FFN_NORM_EXPS, "blk.%d.ffn_norm_exps" },
- { LLM_TENSOR_FFN_GATE_EXPS, "blk.%d.ffn_gate_exps" },
- { LLM_TENSOR_FFN_DOWN_EXPS, "blk.%d.ffn_down_exps" },
- { LLM_TENSOR_FFN_UP_EXPS, "blk.%d.ffn_up_exps" },
- },
- },
- {
- LLM_ARCH_DEEPSEEK,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ROPE_FREQS, "rope_freqs" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_ATTN_ROT_EMBD, "blk.%d.attn_rot_embd" },
- { LLM_TENSOR_FFN_GATE_INP, "blk.%d.ffn_gate_inp" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- { LLM_TENSOR_FFN_GATE_EXPS, "blk.%d.ffn_gate_exps" },
- { LLM_TENSOR_FFN_DOWN_EXPS, "blk.%d.ffn_down_exps" },
- { LLM_TENSOR_FFN_UP_EXPS, "blk.%d.ffn_up_exps" },
- { LLM_TENSOR_FFN_GATE_INP_SHEXP, "blk.%d.ffn_gate_inp_shexp" },
- { LLM_TENSOR_FFN_GATE_SHEXP, "blk.%d.ffn_gate_shexp" },
- { LLM_TENSOR_FFN_DOWN_SHEXP, "blk.%d.ffn_down_shexp" },
- { LLM_TENSOR_FFN_UP_SHEXP, "blk.%d.ffn_up_shexp" },
- },
- },
- {
- LLM_ARCH_DEEPSEEK2,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q_A_NORM, "blk.%d.attn_q_a_norm" },
- { LLM_TENSOR_ATTN_KV_A_NORM, "blk.%d.attn_kv_a_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_Q_A, "blk.%d.attn_q_a" },
- { LLM_TENSOR_ATTN_Q_B, "blk.%d.attn_q_b" },
- { LLM_TENSOR_ATTN_KV_A_MQA, "blk.%d.attn_kv_a_mqa" },
- { LLM_TENSOR_ATTN_KV_B, "blk.%d.attn_kv_b" },
- { LLM_TENSOR_ATTN_K_B, "blk.%d.attn_k_b" },
- { LLM_TENSOR_ATTN_V_B, "blk.%d.attn_v_b" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_GATE_INP, "blk.%d.ffn_gate_inp" },
- { LLM_TENSOR_FFN_GATE_EXPS, "blk.%d.ffn_gate_exps" },
- { LLM_TENSOR_FFN_DOWN_EXPS, "blk.%d.ffn_down_exps" },
- { LLM_TENSOR_FFN_UP_EXPS, "blk.%d.ffn_up_exps" },
- { LLM_TENSOR_FFN_GATE_INP_SHEXP, "blk.%d.ffn_gate_inp_shexp" },
- { LLM_TENSOR_FFN_GATE_SHEXP, "blk.%d.ffn_gate_shexp" },
- { LLM_TENSOR_FFN_DOWN_SHEXP, "blk.%d.ffn_down_shexp" },
- { LLM_TENSOR_FFN_UP_SHEXP, "blk.%d.ffn_up_shexp" },
- { LLM_TENSOR_FFN_EXP_PROBS_B, "blk.%d.exp_probs_b" },
- },
- },
- {
- LLM_ARCH_PLM,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_KV_A_MQA, "blk.%d.attn_kv_a_mqa" },
- { LLM_TENSOR_ATTN_KV_A_NORM, "blk.%d.attn_kv_a_norm" },
- { LLM_TENSOR_ATTN_KV_B, "blk.%d.attn_kv_b" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- },
- },
- {
- LLM_ARCH_CHATGLM,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_ROPE_FREQS, "rope_freqs" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_QKV, "blk.%d.attn_qkv" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- },
- },
- {
- LLM_ARCH_GLM4,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_ROPE_FREQS, "rope_freqs" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_ATTN_POST_NORM, "blk.%d.post_attention_norm" },
- { LLM_TENSOR_FFN_POST_NORM, "blk.%d.post_ffw_norm" },
- },
- },
- {
- LLM_ARCH_GLM4_MOE,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_POST_NORM, "blk.%d.post_attention_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_ATTN_Q_NORM, "blk.%d.attn_q_norm" },
- { LLM_TENSOR_ATTN_K_NORM, "blk.%d.attn_k_norm" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- { LLM_TENSOR_FFN_GATE_INP, "blk.%d.ffn_gate_inp" },
- { LLM_TENSOR_FFN_GATE_EXPS, "blk.%d.ffn_gate_exps" },
- { LLM_TENSOR_FFN_DOWN_EXPS, "blk.%d.ffn_down_exps" },
- { LLM_TENSOR_FFN_UP_EXPS, "blk.%d.ffn_up_exps" },
- { LLM_TENSOR_FFN_GATE_SHEXP, "blk.%d.ffn_gate_shexp" },
- { LLM_TENSOR_FFN_DOWN_SHEXP, "blk.%d.ffn_down_shexp" },
- { LLM_TENSOR_FFN_UP_SHEXP, "blk.%d.ffn_up_shexp" },
- { LLM_TENSOR_FFN_EXP_PROBS_B, "blk.%d.exp_probs_b" },
- // NextN/MTP tensors - preserved but unused (in final layer, dynamic layer number)
- { LLM_TENSOR_NEXTN_EH_PROJ, "blk.%d.nextn.eh_proj" },
- { LLM_TENSOR_NEXTN_EMBED_TOKENS, "blk.%d.nextn.embed_tokens" },
- { LLM_TENSOR_NEXTN_ENORM, "blk.%d.nextn.enorm" },
- { LLM_TENSOR_NEXTN_HNORM, "blk.%d.nextn.hnorm" },
- { LLM_TENSOR_NEXTN_SHARED_HEAD_HEAD, "blk.%d.nextn.shared_head_head" },
- { LLM_TENSOR_NEXTN_SHARED_HEAD_NORM, "blk.%d.nextn.shared_head_norm" },
- },
- },
- {
- LLM_ARCH_BITNET,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_SUB_NORM, "blk.%d.attn_sub_norm" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_SUB_NORM, "blk.%d.ffn_sub_norm" },
- },
- },
- {
- LLM_ARCH_T5,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_DEC_OUTPUT_NORM, "dec.output_norm" },
- { LLM_TENSOR_DEC_ATTN_NORM, "dec.blk.%d.attn_norm" },
- { LLM_TENSOR_DEC_ATTN_Q, "dec.blk.%d.attn_q" },
- { LLM_TENSOR_DEC_ATTN_K, "dec.blk.%d.attn_k" },
- { LLM_TENSOR_DEC_ATTN_V, "dec.blk.%d.attn_v" },
- { LLM_TENSOR_DEC_ATTN_OUT, "dec.blk.%d.attn_o" },
- { LLM_TENSOR_DEC_ATTN_REL_B, "dec.blk.%d.attn_rel_b" },
- { LLM_TENSOR_DEC_CROSS_ATTN_NORM, "dec.blk.%d.cross_attn_norm" },
- { LLM_TENSOR_DEC_CROSS_ATTN_Q, "dec.blk.%d.cross_attn_q" },
- { LLM_TENSOR_DEC_CROSS_ATTN_K, "dec.blk.%d.cross_attn_k" },
- { LLM_TENSOR_DEC_CROSS_ATTN_V, "dec.blk.%d.cross_attn_v" },
- { LLM_TENSOR_DEC_CROSS_ATTN_OUT, "dec.blk.%d.cross_attn_o" },
- { LLM_TENSOR_DEC_CROSS_ATTN_REL_B, "dec.blk.%d.cross_attn_rel_b" },
- { LLM_TENSOR_DEC_FFN_NORM, "dec.blk.%d.ffn_norm" },
- { LLM_TENSOR_DEC_FFN_GATE, "dec.blk.%d.ffn_gate" },
- { LLM_TENSOR_DEC_FFN_DOWN, "dec.blk.%d.ffn_down" },
- { LLM_TENSOR_DEC_FFN_UP, "dec.blk.%d.ffn_up" },
- { LLM_TENSOR_ENC_OUTPUT_NORM, "enc.output_norm" },
- { LLM_TENSOR_ENC_ATTN_NORM, "enc.blk.%d.attn_norm" },
- { LLM_TENSOR_ENC_ATTN_Q, "enc.blk.%d.attn_q" },
- { LLM_TENSOR_ENC_ATTN_K, "enc.blk.%d.attn_k" },
- { LLM_TENSOR_ENC_ATTN_V, "enc.blk.%d.attn_v" },
- { LLM_TENSOR_ENC_ATTN_OUT, "enc.blk.%d.attn_o" },
- { LLM_TENSOR_ENC_ATTN_REL_B, "enc.blk.%d.attn_rel_b" },
- { LLM_TENSOR_ENC_FFN_NORM, "enc.blk.%d.ffn_norm" },
- { LLM_TENSOR_ENC_FFN_GATE, "enc.blk.%d.ffn_gate" },
- { LLM_TENSOR_ENC_FFN_DOWN, "enc.blk.%d.ffn_down" },
- { LLM_TENSOR_ENC_FFN_UP, "enc.blk.%d.ffn_up" },
- },
- },
- {
- LLM_ARCH_T5ENCODER,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ENC_OUTPUT_NORM, "enc.output_norm" },
- { LLM_TENSOR_ENC_ATTN_NORM, "enc.blk.%d.attn_norm" },
- { LLM_TENSOR_ENC_ATTN_Q, "enc.blk.%d.attn_q" },
- { LLM_TENSOR_ENC_ATTN_K, "enc.blk.%d.attn_k" },
- { LLM_TENSOR_ENC_ATTN_V, "enc.blk.%d.attn_v" },
- { LLM_TENSOR_ENC_ATTN_OUT, "enc.blk.%d.attn_o" },
- { LLM_TENSOR_ENC_ATTN_REL_B, "enc.blk.%d.attn_rel_b" },
- { LLM_TENSOR_ENC_FFN_NORM, "enc.blk.%d.ffn_norm" },
- { LLM_TENSOR_ENC_FFN_GATE, "enc.blk.%d.ffn_gate" },
- { LLM_TENSOR_ENC_FFN_DOWN, "enc.blk.%d.ffn_down" },
- { LLM_TENSOR_ENC_FFN_UP, "enc.blk.%d.ffn_up" },
- },
- },
- {
- LLM_ARCH_JAIS,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_QKV, "blk.%d.attn_qkv" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- },
- },
- {
- LLM_ARCH_NEMOTRON,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ROPE_FREQS, "rope_freqs" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_ATTN_ROT_EMBD, "blk.%d.attn_rot_embd" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- },
- },
- {
- LLM_ARCH_NEMOTRON_H,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- // mamba(2) ssm layers
- { LLM_TENSOR_SSM_IN, "blk.%d.ssm_in" },
- { LLM_TENSOR_SSM_CONV1D, "blk.%d.ssm_conv1d" },
- { LLM_TENSOR_SSM_DT, "blk.%d.ssm_dt" },
- { LLM_TENSOR_SSM_A, "blk.%d.ssm_a" },
- { LLM_TENSOR_SSM_D, "blk.%d.ssm_d" },
- { LLM_TENSOR_SSM_NORM, "blk.%d.ssm_norm" },
- { LLM_TENSOR_SSM_OUT, "blk.%d.ssm_out" },
- // attention layers
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- // dense FFN
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- },
- },
- {
- LLM_ARCH_EXAONE,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ROPE_FREQS, "rope_freqs" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_ATTN_ROT_EMBD, "blk.%d.attn_rot_embd" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- },
- },
- {
- LLM_ARCH_EXAONE4,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ROPE_FREQS, "rope_freqs" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_Q_NORM, "blk.%d.attn_q_norm" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_K_NORM, "blk.%d.attn_k_norm" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_ATTN_POST_NORM, "blk.%d.post_attention_norm" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- { LLM_TENSOR_FFN_POST_NORM, "blk.%d.post_ffw_norm" },
- }
- },
- {
- LLM_ARCH_RWKV6,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_TOKEN_EMBD_NORM, "token_embd_norm" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_NORM_2, "blk.%d.attn_norm_2" },
- { LLM_TENSOR_TIME_MIX_W1, "blk.%d.time_mix_w1" },
- { LLM_TENSOR_TIME_MIX_W2, "blk.%d.time_mix_w2" },
- { LLM_TENSOR_TIME_MIX_LERP_X, "blk.%d.time_mix_lerp_x" },
- { LLM_TENSOR_TIME_MIX_LERP_W, "blk.%d.time_mix_lerp_w" },
- { LLM_TENSOR_TIME_MIX_LERP_K, "blk.%d.time_mix_lerp_k" },
- { LLM_TENSOR_TIME_MIX_LERP_V, "blk.%d.time_mix_lerp_v" },
- { LLM_TENSOR_TIME_MIX_LERP_R, "blk.%d.time_mix_lerp_r" },
- { LLM_TENSOR_TIME_MIX_LERP_G, "blk.%d.time_mix_lerp_g" },
- { LLM_TENSOR_TIME_MIX_LERP_FUSED, "blk.%d.time_mix_lerp_fused" },
- { LLM_TENSOR_TIME_MIX_FIRST, "blk.%d.time_mix_first" },
- { LLM_TENSOR_TIME_MIX_DECAY, "blk.%d.time_mix_decay" },
- { LLM_TENSOR_TIME_MIX_DECAY_W1, "blk.%d.time_mix_decay_w1" },
- { LLM_TENSOR_TIME_MIX_DECAY_W2, "blk.%d.time_mix_decay_w2" },
- { LLM_TENSOR_TIME_MIX_KEY, "blk.%d.time_mix_key" },
- { LLM_TENSOR_TIME_MIX_VALUE, "blk.%d.time_mix_value" },
- { LLM_TENSOR_TIME_MIX_RECEPTANCE, "blk.%d.time_mix_receptance" },
- { LLM_TENSOR_TIME_MIX_GATE, "blk.%d.time_mix_gate" },
- { LLM_TENSOR_TIME_MIX_LN, "blk.%d.time_mix_ln" },
- { LLM_TENSOR_TIME_MIX_OUTPUT, "blk.%d.time_mix_output" },
- { LLM_TENSOR_CHANNEL_MIX_LERP_K, "blk.%d.channel_mix_lerp_k" },
- { LLM_TENSOR_CHANNEL_MIX_LERP_R, "blk.%d.channel_mix_lerp_r" },
- { LLM_TENSOR_CHANNEL_MIX_KEY, "blk.%d.channel_mix_key" },
- { LLM_TENSOR_CHANNEL_MIX_VALUE, "blk.%d.channel_mix_value" },
- { LLM_TENSOR_CHANNEL_MIX_RECEPTANCE, "blk.%d.channel_mix_receptance" },
- },
- },
- {
- LLM_ARCH_RWKV6QWEN2,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_TIME_MIX_W1, "blk.%d.time_mix_w1" },
- { LLM_TENSOR_TIME_MIX_W2, "blk.%d.time_mix_w2" },
- { LLM_TENSOR_TIME_MIX_LERP_X, "blk.%d.time_mix_lerp_x" },
- { LLM_TENSOR_TIME_MIX_LERP_FUSED, "blk.%d.time_mix_lerp_fused" },
- { LLM_TENSOR_TIME_MIX_FIRST, "blk.%d.time_mix_first" },
- { LLM_TENSOR_TIME_MIX_DECAY, "blk.%d.time_mix_decay" },
- { LLM_TENSOR_TIME_MIX_DECAY_W1, "blk.%d.time_mix_decay_w1" },
- { LLM_TENSOR_TIME_MIX_DECAY_W2, "blk.%d.time_mix_decay_w2" },
- { LLM_TENSOR_TIME_MIX_KEY, "blk.%d.time_mix_key" },
- { LLM_TENSOR_TIME_MIX_VALUE, "blk.%d.time_mix_value" },
- { LLM_TENSOR_TIME_MIX_RECEPTANCE, "blk.%d.time_mix_receptance" },
- { LLM_TENSOR_TIME_MIX_GATE, "blk.%d.time_mix_gate" },
- { LLM_TENSOR_TIME_MIX_OUTPUT, "blk.%d.time_mix_output" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- },
- },
- {
- LLM_ARCH_RWKV7,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_TOKEN_EMBD_NORM, "token_embd_norm" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_NORM_2, "blk.%d.attn_norm_2" },
- { LLM_TENSOR_TIME_MIX_W0, "blk.%d.time_mix_w0" },
- { LLM_TENSOR_TIME_MIX_W1, "blk.%d.time_mix_w1" },
- { LLM_TENSOR_TIME_MIX_W2, "blk.%d.time_mix_w2" },
- { LLM_TENSOR_TIME_MIX_A0, "blk.%d.time_mix_a0" },
- { LLM_TENSOR_TIME_MIX_A1, "blk.%d.time_mix_a1" },
- { LLM_TENSOR_TIME_MIX_A2, "blk.%d.time_mix_a2" },
- { LLM_TENSOR_TIME_MIX_V0, "blk.%d.time_mix_v0" },
- { LLM_TENSOR_TIME_MIX_V1, "blk.%d.time_mix_v1" },
- { LLM_TENSOR_TIME_MIX_V2, "blk.%d.time_mix_v2" },
- { LLM_TENSOR_TIME_MIX_G1, "blk.%d.time_mix_g1" },
- { LLM_TENSOR_TIME_MIX_G2, "blk.%d.time_mix_g2" },
- { LLM_TENSOR_TIME_MIX_K_K, "blk.%d.time_mix_k_k" },
- { LLM_TENSOR_TIME_MIX_K_A, "blk.%d.time_mix_k_a" },
- { LLM_TENSOR_TIME_MIX_R_K, "blk.%d.time_mix_r_k" },
- { LLM_TENSOR_TIME_MIX_LERP_FUSED, "blk.%d.time_mix_lerp_fused" },
- { LLM_TENSOR_TIME_MIX_KEY, "blk.%d.time_mix_key" },
- { LLM_TENSOR_TIME_MIX_VALUE, "blk.%d.time_mix_value" },
- { LLM_TENSOR_TIME_MIX_RECEPTANCE, "blk.%d.time_mix_receptance" },
- { LLM_TENSOR_TIME_MIX_LN, "blk.%d.time_mix_ln" },
- { LLM_TENSOR_TIME_MIX_OUTPUT, "blk.%d.time_mix_output" },
- { LLM_TENSOR_CHANNEL_MIX_LERP_K, "blk.%d.channel_mix_lerp_k" },
- { LLM_TENSOR_CHANNEL_MIX_KEY, "blk.%d.channel_mix_key" },
- { LLM_TENSOR_CHANNEL_MIX_VALUE, "blk.%d.channel_mix_value" },
- },
- },
- {
- LLM_ARCH_ARWKV7,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_TOKEN_EMBD_NORM, "token_embd_norm" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_TIME_MIX_W0, "blk.%d.time_mix_w0" },
- { LLM_TENSOR_TIME_MIX_W1, "blk.%d.time_mix_w1" },
- { LLM_TENSOR_TIME_MIX_W2, "blk.%d.time_mix_w2" },
- { LLM_TENSOR_TIME_MIX_A0, "blk.%d.time_mix_a0" },
- { LLM_TENSOR_TIME_MIX_A1, "blk.%d.time_mix_a1" },
- { LLM_TENSOR_TIME_MIX_A2, "blk.%d.time_mix_a2" },
- { LLM_TENSOR_TIME_MIX_V0, "blk.%d.time_mix_v0" },
- { LLM_TENSOR_TIME_MIX_V1, "blk.%d.time_mix_v1" },
- { LLM_TENSOR_TIME_MIX_V2, "blk.%d.time_mix_v2" },
- { LLM_TENSOR_TIME_MIX_G1, "blk.%d.time_mix_g1" },
- { LLM_TENSOR_TIME_MIX_G2, "blk.%d.time_mix_g2" },
- { LLM_TENSOR_TIME_MIX_K_K, "blk.%d.time_mix_k_k" },
- { LLM_TENSOR_TIME_MIX_K_A, "blk.%d.time_mix_k_a" },
- { LLM_TENSOR_TIME_MIX_R_K, "blk.%d.time_mix_r_k" },
- { LLM_TENSOR_TIME_MIX_LERP_FUSED, "blk.%d.time_mix_lerp_fused" },
- { LLM_TENSOR_TIME_MIX_KEY, "blk.%d.time_mix_key" },
- { LLM_TENSOR_TIME_MIX_VALUE, "blk.%d.time_mix_value" },
- { LLM_TENSOR_TIME_MIX_RECEPTANCE, "blk.%d.time_mix_receptance" },
- { LLM_TENSOR_TIME_MIX_LN, "blk.%d.time_mix_ln" },
- { LLM_TENSOR_TIME_MIX_OUTPUT, "blk.%d.time_mix_output" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- },
- },
- {
- LLM_ARCH_GRANITE,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- },
- },
- {
- LLM_ARCH_GRANITE_MOE,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_GATE_INP, "blk.%d.ffn_gate_inp" },
- { LLM_TENSOR_FFN_GATE_EXPS, "blk.%d.ffn_gate_exps" },
- { LLM_TENSOR_FFN_DOWN_EXPS, "blk.%d.ffn_down_exps" },
- { LLM_TENSOR_FFN_UP_EXPS, "blk.%d.ffn_up_exps" },
- { LLM_TENSOR_FFN_GATE_SHEXP, "blk.%d.ffn_gate_shexp" },
- { LLM_TENSOR_FFN_DOWN_SHEXP, "blk.%d.ffn_down_shexp" },
- { LLM_TENSOR_FFN_UP_SHEXP, "blk.%d.ffn_up_shexp" },
- },
- },
- {
- LLM_ARCH_GRANITE_HYBRID,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- // mamba(2) ssm layers
- { LLM_TENSOR_SSM_IN, "blk.%d.ssm_in" },
- { LLM_TENSOR_SSM_CONV1D, "blk.%d.ssm_conv1d" },
- { LLM_TENSOR_SSM_DT, "blk.%d.ssm_dt" },
- { LLM_TENSOR_SSM_A, "blk.%d.ssm_a" },
- { LLM_TENSOR_SSM_D, "blk.%d.ssm_d" },
- { LLM_TENSOR_SSM_NORM, "blk.%d.ssm_norm" },
- { LLM_TENSOR_SSM_OUT, "blk.%d.ssm_out" },
- // attention layers
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- // dense FFN
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- // moe FFN
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_GATE_INP, "blk.%d.ffn_gate_inp" },
- { LLM_TENSOR_FFN_GATE_EXPS, "blk.%d.ffn_gate_exps" },
- { LLM_TENSOR_FFN_DOWN_EXPS, "blk.%d.ffn_down_exps" },
- { LLM_TENSOR_FFN_UP_EXPS, "blk.%d.ffn_up_exps" },
- // shared expert
- { LLM_TENSOR_FFN_GATE_SHEXP, "blk.%d.ffn_gate_shexp" },
- { LLM_TENSOR_FFN_DOWN_SHEXP, "blk.%d.ffn_down_shexp" },
- { LLM_TENSOR_FFN_UP_SHEXP, "blk.%d.ffn_up_shexp" },
- },
- },
- {
- LLM_ARCH_CHAMELEON,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- { LLM_TENSOR_ATTN_Q_NORM, "blk.%d.attn_q_norm" },
- { LLM_TENSOR_ATTN_K_NORM, "blk.%d.attn_k_norm" },
- },
- },
- {
- LLM_ARCH_WAVTOKENIZER_DEC,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_TOKEN_EMBD_NORM, "token_embd_norm" },
- { LLM_TENSOR_CONV1D, "conv1d" },
- { LLM_TENSOR_CONVNEXT_DW, "convnext.%d.dw" },
- { LLM_TENSOR_CONVNEXT_NORM, "convnext.%d.norm" },
- { LLM_TENSOR_CONVNEXT_PW1, "convnext.%d.pw1" },
- { LLM_TENSOR_CONVNEXT_PW2, "convnext.%d.pw2" },
- { LLM_TENSOR_CONVNEXT_GAMMA, "convnext.%d.gamma" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_POS_NET_CONV1, "posnet.%d.conv1" },
- { LLM_TENSOR_POS_NET_CONV2, "posnet.%d.conv2" },
- { LLM_TENSOR_POS_NET_NORM, "posnet.%d.norm" },
- { LLM_TENSOR_POS_NET_NORM1, "posnet.%d.norm1" },
- { LLM_TENSOR_POS_NET_NORM2, "posnet.%d.norm2" },
- { LLM_TENSOR_POS_NET_ATTN_NORM, "posnet.%d.attn_norm" },
- { LLM_TENSOR_POS_NET_ATTN_Q, "posnet.%d.attn_q" },
- { LLM_TENSOR_POS_NET_ATTN_K, "posnet.%d.attn_k" },
- { LLM_TENSOR_POS_NET_ATTN_V, "posnet.%d.attn_v" },
- { LLM_TENSOR_POS_NET_ATTN_OUT, "posnet.%d.attn_output" },
- },
- },
- {
- LLM_ARCH_BAILINGMOE,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ROPE_FREQS, "rope_freqs" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_FFN_GATE_INP, "blk.%d.ffn_gate_inp" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_GATE_EXPS, "blk.%d.ffn_gate_exps" },
- { LLM_TENSOR_FFN_DOWN_EXPS, "blk.%d.ffn_down_exps" },
- { LLM_TENSOR_FFN_UP_EXPS, "blk.%d.ffn_up_exps" },
- { LLM_TENSOR_FFN_GATE_INP_SHEXP, "blk.%d.ffn_gate_inp_shexp" },
- { LLM_TENSOR_FFN_GATE_SHEXP, "blk.%d.ffn_gate_shexp" },
- { LLM_TENSOR_FFN_DOWN_SHEXP, "blk.%d.ffn_down_shexp" },
- { LLM_TENSOR_FFN_UP_SHEXP, "blk.%d.ffn_up_shexp" },
- },
- },
- {
- LLM_ARCH_BAILINGMOE2,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q_NORM, "blk.%d.attn_q_norm" },
- { LLM_TENSOR_ATTN_K_NORM, "blk.%d.attn_k_norm" },
- { LLM_TENSOR_ATTN_QKV, "blk.%d.attn_qkv" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_FFN_GATE_INP, "blk.%d.ffn_gate_inp" },
- { LLM_TENSOR_FFN_EXP_PROBS_B, "blk.%d.exp_probs_b" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- { LLM_TENSOR_FFN_GATE_EXPS, "blk.%d.ffn_gate_exps" },
- { LLM_TENSOR_FFN_DOWN_EXPS, "blk.%d.ffn_down_exps" },
- { LLM_TENSOR_FFN_UP_EXPS, "blk.%d.ffn_up_exps" },
- { LLM_TENSOR_FFN_GATE_SHEXP, "blk.%d.ffn_gate_shexp" },
- { LLM_TENSOR_FFN_DOWN_SHEXP, "blk.%d.ffn_down_shexp" },
- { LLM_TENSOR_FFN_UP_SHEXP, "blk.%d.ffn_up_shexp" },
- { LLM_TENSOR_NEXTN_EH_PROJ, "blk.%d.nextn.eh_proj" },
- { LLM_TENSOR_NEXTN_EMBED_TOKENS, "blk.%d.nextn.embed_tokens" },
- { LLM_TENSOR_NEXTN_ENORM, "blk.%d.nextn.enorm" },
- { LLM_TENSOR_NEXTN_HNORM, "blk.%d.nextn.hnorm" },
- { LLM_TENSOR_NEXTN_SHARED_HEAD_HEAD, "blk.%d.nextn.shared_head_head" },
- { LLM_TENSOR_NEXTN_SHARED_HEAD_NORM, "blk.%d.nextn.shared_head_norm" },
- { LLM_TENSOR_LAYER_OUT_NORM, "blk.%d.layer_output_norm" },
- },
- },
- {
- LLM_ARCH_DOTS1,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_Q_NORM, "blk.%d.attn_q_norm" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_K_NORM, "blk.%d.attn_k_norm" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_GATE_INP, "blk.%d.ffn_gate_inp" },
- { LLM_TENSOR_FFN_GATE_EXPS, "blk.%d.ffn_gate_exps" },
- { LLM_TENSOR_FFN_DOWN_EXPS, "blk.%d.ffn_down_exps" },
- { LLM_TENSOR_FFN_UP_EXPS, "blk.%d.ffn_up_exps" },
- { LLM_TENSOR_FFN_GATE_INP_SHEXP, "blk.%d.ffn_gate_inp_shexp" },
- { LLM_TENSOR_FFN_GATE_SHEXP, "blk.%d.ffn_gate_shexp" },
- { LLM_TENSOR_FFN_DOWN_SHEXP, "blk.%d.ffn_down_shexp" },
- { LLM_TENSOR_FFN_UP_SHEXP, "blk.%d.ffn_up_shexp" },
- { LLM_TENSOR_FFN_EXP_PROBS_B, "blk.%d.exp_probs_b" },
- }
- },
- {
- LLM_ARCH_ERNIE4_5,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- },
- },
- {
- LLM_ARCH_ERNIE4_5_MOE,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- { LLM_TENSOR_FFN_GATE_INP, "blk.%d.ffn_gate_inp" },
- { LLM_TENSOR_FFN_GATE_SHEXP, "blk.%d.ffn_gate_shexp" },
- { LLM_TENSOR_FFN_DOWN_SHEXP, "blk.%d.ffn_down_shexp" },
- { LLM_TENSOR_FFN_UP_SHEXP, "blk.%d.ffn_up_shexp" },
- { LLM_TENSOR_FFN_GATE_EXPS, "blk.%d.ffn_gate_exps" },
- { LLM_TENSOR_FFN_DOWN_EXPS, "blk.%d.ffn_down_exps" },
- { LLM_TENSOR_FFN_UP_EXPS, "blk.%d.ffn_up_exps" },
- { LLM_TENSOR_FFN_EXP_PROBS_B, "blk.%d.exp_probs_b" },
- },
- },
- {
- LLM_ARCH_HUNYUAN_MOE,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_Q_NORM, "blk.%d.attn_q_norm" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_K_NORM, "blk.%d.attn_k_norm" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_FFN_GATE_INP, "blk.%d.ffn_gate_inp" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_GATE_SHEXP, "blk.%d.ffn_gate_shexp" },
- { LLM_TENSOR_FFN_DOWN_SHEXP, "blk.%d.ffn_down_shexp" },
- { LLM_TENSOR_FFN_UP_SHEXP, "blk.%d.ffn_up_shexp" },
- { LLM_TENSOR_FFN_GATE_EXPS, "blk.%d.ffn_gate_exps" },
- { LLM_TENSOR_FFN_DOWN_EXPS, "blk.%d.ffn_down_exps" },
- { LLM_TENSOR_FFN_UP_EXPS, "blk.%d.ffn_up_exps" },
- },
- },
- {
- LLM_ARCH_HUNYUAN_DENSE,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_Q_NORM, "blk.%d.attn_q_norm" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_K_NORM, "blk.%d.attn_k_norm" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
-
- },
- },
- {
- LLM_ARCH_SMOLLM3,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- },
- },
- {
- LLM_ARCH_OPENAI_MOE,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_POST_NORM, "blk.%d.post_attention_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_ATTN_SINKS, "blk.%d.attn_sinks" },
- { LLM_TENSOR_FFN_GATE_INP, "blk.%d.ffn_gate_inp" },
- { LLM_TENSOR_FFN_GATE_EXPS, "blk.%d.ffn_gate_exps" },
- { LLM_TENSOR_FFN_DOWN_EXPS, "blk.%d.ffn_down_exps" },
- { LLM_TENSOR_FFN_UP_EXPS, "blk.%d.ffn_up_exps" },
- },
- },
- {
- LLM_ARCH_LFM2,
- {
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_ATTN_K_NORM, "blk.%d.attn_k_norm" },
- { LLM_TENSOR_ATTN_Q_NORM, "blk.%d.attn_q_norm" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- { LLM_TENSOR_SHORTCONV_CONV, "blk.%d.shortconv.conv" },
- { LLM_TENSOR_SHORTCONV_INPROJ, "blk.%d.shortconv.in_proj" },
- { LLM_TENSOR_SHORTCONV_OUTPROJ, "blk.%d.shortconv.out_proj" },
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "token_embd_norm" }, // note: wrong tensor name
- { LLM_TENSOR_OUTPUT, "output" },
- }
- },
- {
- LLM_ARCH_LFM2MOE,
- {
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_ATTN_K_NORM, "blk.%d.attn_k_norm" },
- { LLM_TENSOR_ATTN_Q_NORM, "blk.%d.attn_q_norm" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- { LLM_TENSOR_SHORTCONV_CONV, "blk.%d.shortconv.conv" },
- { LLM_TENSOR_SHORTCONV_INPROJ, "blk.%d.shortconv.in_proj" },
- { LLM_TENSOR_SHORTCONV_OUTPROJ, "blk.%d.shortconv.out_proj" },
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "token_embd_norm" }, // note: wrong tensor name
- { LLM_TENSOR_FFN_GATE_INP, "blk.%d.ffn_gate_inp" },
- { LLM_TENSOR_FFN_GATE_EXPS, "blk.%d.ffn_gate_exps" },
- { LLM_TENSOR_FFN_DOWN_EXPS, "blk.%d.ffn_down_exps" },
- { LLM_TENSOR_FFN_UP_EXPS, "blk.%d.ffn_up_exps" },
- { LLM_TENSOR_FFN_EXP_PROBS_B, "blk.%d.exp_probs_b" },
- }
- },
- {
- LLM_ARCH_SMALLTHINKER,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- { LLM_TENSOR_FFN_GATE_INP, "blk.%d.ffn_gate_inp" },
- { LLM_TENSOR_FFN_GATE_EXPS, "blk.%d.ffn_gate_exps" },
- { LLM_TENSOR_FFN_DOWN_EXPS, "blk.%d.ffn_down_exps" },
- { LLM_TENSOR_FFN_UP_EXPS, "blk.%d.ffn_up_exps" }
- },
- },
- {
- LLM_ARCH_APERTUS,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ROPE_FREQS, "rope_freqs" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_ATTN_Q_NORM, "blk.%d.attn_q_norm" },
- { LLM_TENSOR_ATTN_K_NORM, "blk.%d.attn_k_norm" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- },
- },
- {
- LLM_ARCH_DREAM,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- },
- },
- {
- LLM_ARCH_LLADA,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- },
- },
- {
- LLM_ARCH_LLADA_MOE,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_Q_NORM, "blk.%d.attn_q_norm" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_K_NORM, "blk.%d.attn_k_norm" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_GATE_INP, "blk.%d.ffn_gate_inp" },
- { LLM_TENSOR_FFN_GATE_EXPS, "blk.%d.ffn_gate_exps" },
- { LLM_TENSOR_FFN_DOWN_EXPS, "blk.%d.ffn_down_exps" },
- { LLM_TENSOR_FFN_UP_EXPS, "blk.%d.ffn_up_exps" },
- },
- },
- {
- LLM_ARCH_SEED_OSS,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_ATTN_POST_NORM, "blk.%d.post_attention_norm" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- },
- },
- {
- LLM_ARCH_GROVEMOE,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_Q_NORM, "blk.%d.attn_q_norm" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_K_NORM, "blk.%d.attn_k_norm" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_GATE_INP, "blk.%d.ffn_gate_inp" },
- { LLM_TENSOR_FFN_GATE_EXPS, "blk.%d.ffn_gate_exps" },
- { LLM_TENSOR_FFN_DOWN_EXPS, "blk.%d.ffn_down_exps" },
- { LLM_TENSOR_FFN_UP_EXPS, "blk.%d.ffn_up_exps" },
- { LLM_TENSOR_FFN_GATE_CHEXPS, "blk.%d.ffn_gate_chexps" },
- { LLM_TENSOR_FFN_DOWN_CHEXPS, "blk.%d.ffn_down_chexps" },
- { LLM_TENSOR_FFN_UP_CHEXPS, "blk.%d.ffn_up_chexps" },
- },
- },
- {
- LLM_ARCH_MINIMAX_M2,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_ATTN_Q_NORM, "blk.%d.attn_q_norm" },
- { LLM_TENSOR_ATTN_K_NORM, "blk.%d.attn_k_norm" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_GATE_INP, "blk.%d.ffn_gate_inp" },
- { LLM_TENSOR_FFN_GATE_EXPS, "blk.%d.ffn_gate_exps" },
- { LLM_TENSOR_FFN_DOWN_EXPS, "blk.%d.ffn_down_exps" },
- { LLM_TENSOR_FFN_UP_EXPS, "blk.%d.ffn_up_exps" },
- { LLM_TENSOR_FFN_EXP_PROBS_B, "blk.%d.exp_probs_b" },
- },
- },
- {
- LLM_ARCH_PANGU_EMBED,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- },
- },
- {
- LLM_ARCH_COGVLM,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_QKV, "blk.%d.attn_qkv" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- { LLM_TENSOR_VISEXP_ATTN_QKV, "blk.%d.vis_attn_qkv" },
- { LLM_TENSOR_VISEXP_ATTN_OUT, "blk.%d.vis_attn_output" },
- { LLM_TENSOR_VISEXP_FFN_GATE, "blk.%d.vis_gate" },
- { LLM_TENSOR_VISEXP_FFN_DOWN, "blk.%d.vis_down" },
- { LLM_TENSOR_VISEXP_FFN_UP, "blk.%d.vis_up" },
- },
- },
- {
- LLM_ARCH_RND1,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_Q_NORM, "blk.%d.attn_q_norm" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_K_NORM, "blk.%d.attn_k_norm" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_GATE_INP, "blk.%d.ffn_gate_inp" },
- { LLM_TENSOR_FFN_GATE_EXPS, "blk.%d.ffn_gate_exps" },
- { LLM_TENSOR_FFN_DOWN_EXPS, "blk.%d.ffn_down_exps" },
- { LLM_TENSOR_FFN_UP_EXPS, "blk.%d.ffn_up_exps" },
- },
- },
- {
- LLM_ARCH_MISTRAL3,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
- { LLM_TENSOR_OUTPUT, "output" },
- { LLM_TENSOR_ROPE_FREQS, "rope_freqs" },
- { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
- { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
- { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
- { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
- { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
- { LLM_TENSOR_ATTN_ROT_EMBD, "blk.%d.attn_rot_embd" },
- { LLM_TENSOR_FFN_GATE_INP, "blk.%d.ffn_gate_inp" },
- { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
- { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
- { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
- { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
- { LLM_TENSOR_FFN_GATE_EXP, "blk.%d.ffn_gate.%d" },
- { LLM_TENSOR_FFN_DOWN_EXP, "blk.%d.ffn_down.%d" },
- { LLM_TENSOR_FFN_UP_EXP, "blk.%d.ffn_up.%d" },
- { LLM_TENSOR_FFN_GATE_EXPS, "blk.%d.ffn_gate_exps" },
- { LLM_TENSOR_FFN_DOWN_EXPS, "blk.%d.ffn_down_exps" },
- { LLM_TENSOR_FFN_UP_EXPS, "blk.%d.ffn_up_exps" },
- },
- },
- {
- LLM_ARCH_UNKNOWN,
- {
- { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
- },
- },
+static const std::map<llm_tensor, const char *> LLM_TENSOR_NAMES = {
+ { LLM_TENSOR_TOKEN_EMBD, "token_embd" },
+ { LLM_TENSOR_OUTPUT_NORM, "output_norm" },
+ { LLM_TENSOR_OUTPUT_NORM_LFM2, "token_embd_norm" }, // fix for wrong tensor name
+ { LLM_TENSOR_OUTPUT, "output" },
+ { LLM_TENSOR_ROPE_FREQS, "rope_freqs" },
+ { LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
+ { LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
+ { LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
+ { LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
+ { LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
+ { LLM_TENSOR_ATTN_ROT_EMBD, "blk.%d.attn_rot_embd" },
+ { LLM_TENSOR_FFN_GATE_INP, "blk.%d.ffn_gate_inp" },
+ { LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
+ { LLM_TENSOR_FFN_GATE, "blk.%d.ffn_gate" },
+ { LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
+ { LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
+ { LLM_TENSOR_FFN_GATE_EXP, "blk.%d.ffn_gate.%d" },
+ { LLM_TENSOR_FFN_DOWN_EXP, "blk.%d.ffn_down.%d" },
+ { LLM_TENSOR_FFN_UP_EXP, "blk.%d.ffn_up.%d" },
+ { LLM_TENSOR_FFN_GATE_EXPS, "blk.%d.ffn_gate_exps" },
+ { LLM_TENSOR_FFN_DOWN_EXPS, "blk.%d.ffn_down_exps" },
+ { LLM_TENSOR_FFN_UP_EXPS, "blk.%d.ffn_up_exps" },
+ { LLM_TENSOR_ATTN_POST_NORM, "blk.%d.post_attention_norm" },
+ { LLM_TENSOR_ATTN_Q_NORM, "blk.%d.attn_q_norm" },
+ { LLM_TENSOR_ATTN_K_NORM, "blk.%d.attn_k_norm" },
+ { LLM_TENSOR_ATTN_GATE, "blk.%d.attn_gate" },
+ { LLM_TENSOR_FFN_POST_NORM, "blk.%d.post_ffw_norm" },
+ { LLM_TENSOR_FFN_GATE_SHEXP, "blk.%d.ffn_gate_shexp" },
+ { LLM_TENSOR_FFN_UP_SHEXP, "blk.%d.ffn_up_shexp" },
+ { LLM_TENSOR_FFN_DOWN_SHEXP, "blk.%d.ffn_down_shexp" },
+ { LLM_TENSOR_FFN_EXP_PROBS_B, "blk.%d.exp_probs_b" },
+ { LLM_TENSOR_ATTN_NORM_2, "blk.%d.attn_norm_2" },
+ { LLM_TENSOR_ATTN_QKV, "blk.%d.attn_qkv" },
+ { LLM_TENSOR_LAYER_OUT_NORM, "blk.%d.layer_output_norm" },
+ { LLM_TENSOR_ATTN_OUT_NORM, "blk.%d.attn_output_norm" },
+ { LLM_TENSOR_POS_EMBD, "position_embd" },
+ { LLM_TENSOR_FFN_ACT, "blk.%d.ffn.act" },
+ { LLM_TENSOR_TOKEN_EMBD_NORM, "token_embd_norm" },
+ { LLM_TENSOR_TOKEN_TYPES, "token_types" },
+ { LLM_TENSOR_CLS, "cls" },
+ { LLM_TENSOR_CLS_OUT, "cls.output" },
+ { LLM_TENSOR_ENC_OUTPUT_NORM, "enc.output_norm" },
+ { LLM_TENSOR_FFN_GATE_INP_SHEXP, "blk.%d.ffn_gate_inp_shexp" },
+ { LLM_TENSOR_SSM_A_NOSCAN, "blk.%d.ssm_a" },
+ { LLM_TENSOR_SSM_CONV1D, "blk.%d.ssm_conv1d" },
+ { LLM_TENSOR_SSM_DT, "blk.%d.ssm_dt" },
+ { LLM_TENSOR_SSM_BETA_ALPHA, "blk.%d.ssm_ba" },
+ { LLM_TENSOR_SSM_IN, "blk.%d.ssm_in" },
+ { LLM_TENSOR_SSM_NORM, "blk.%d.ssm_norm" },
+ { LLM_TENSOR_SSM_OUT, "blk.%d.ssm_out" },
+ { LLM_TENSOR_ROPE_FACTORS_LONG, "rope_factors_long" },
+ { LLM_TENSOR_ROPE_FACTORS_SHORT, "rope_factors_short" },
+ { LLM_TENSOR_SSM_X, "blk.%d.ssm_x" },
+ { LLM_TENSOR_SSM_A, "blk.%d.ssm_a" },
+ { LLM_TENSOR_SSM_D, "blk.%d.ssm_d" },
+ { LLM_TENSOR_SSM_DT_NORM, "blk.%d.ssm_dt_norm" },
+ { LLM_TENSOR_SSM_B_NORM, "blk.%d.ssm_b_norm" },
+ { LLM_TENSOR_SSM_C_NORM, "blk.%d.ssm_c_norm" },
+ { LLM_TENSOR_ATTN_Q_A_NORM, "blk.%d.attn_q_a_norm" },
+ { LLM_TENSOR_ATTN_KV_A_NORM, "blk.%d.attn_kv_a_norm" },
+ { LLM_TENSOR_ATTN_Q_A, "blk.%d.attn_q_a" },
+ { LLM_TENSOR_ATTN_Q_B, "blk.%d.attn_q_b" },
+ { LLM_TENSOR_ATTN_KV_A_MQA, "blk.%d.attn_kv_a_mqa" },
+ { LLM_TENSOR_ATTN_KV_B, "blk.%d.attn_kv_b" },
+ { LLM_TENSOR_PER_LAYER_TOKEN_EMBD, "per_layer_token_embd" },
+ { LLM_TENSOR_PER_LAYER_MODEL_PROJ, "per_layer_model_proj" },
+ { LLM_TENSOR_PER_LAYER_PROJ_NORM, "per_layer_proj_norm" },
+ { LLM_TENSOR_ALTUP_UNEMBD_PROJ, "altup_unembd_proj" },
+ { LLM_TENSOR_ALTUP_PROJ, "altup_proj" },
+ { LLM_TENSOR_PER_LAYER_INP_GATE, "blk.%d.inp_gate" },
+ { LLM_TENSOR_PER_LAYER_PROJ, "blk.%d.proj" },
+ { LLM_TENSOR_PER_LAYER_POST_NORM, "blk.%d.post_norm" },
+ { LLM_TENSOR_ALTUP_CORRECT_COEF, "blk.%d.altup_correct_coef" },
+ { LLM_TENSOR_ALTUP_CORRECT_SCALE, "blk.%d.altup_correct_scale" },
+ { LLM_TENSOR_ALTUP_PREDICT_COEF, "blk.%d.altup_predict_coef" },
+ { LLM_TENSOR_ALTUP_ROUTER, "blk.%d.altup_router" },
+ { LLM_TENSOR_ALTUP_ROUTER_NORM, "blk.%d.altup_router_norm" },
+ { LLM_TENSOR_LAUREL_L, "blk.%d.laurel_l" },
+ { LLM_TENSOR_LAUREL_R, "blk.%d.laurel_r" },
+ { LLM_TENSOR_LAUREL_POST_NORM, "blk.%d.laurel_post_norm" },
+ { LLM_TENSOR_DENSE_2_OUT, "dense_2" },
+ { LLM_TENSOR_DENSE_3_OUT, "dense_3" },
+ { LLM_TENSOR_FFN_NORM_EXPS, "blk.%d.ffn_norm_exps" },
+ { LLM_TENSOR_ATTN_K_B, "blk.%d.attn_k_b" },
+ { LLM_TENSOR_ATTN_V_B, "blk.%d.attn_v_b" },
+ { LLM_TENSOR_NEXTN_EH_PROJ, "blk.%d.nextn.eh_proj" },
+ { LLM_TENSOR_NEXTN_EMBED_TOKENS, "blk.%d.nextn.embed_tokens" },
+ { LLM_TENSOR_NEXTN_ENORM, "blk.%d.nextn.enorm" },
+ { LLM_TENSOR_NEXTN_HNORM, "blk.%d.nextn.hnorm" },
+ { LLM_TENSOR_NEXTN_SHARED_HEAD_HEAD, "blk.%d.nextn.shared_head_head" },
+ { LLM_TENSOR_NEXTN_SHARED_HEAD_NORM, "blk.%d.nextn.shared_head_norm" },
+ { LLM_TENSOR_ATTN_SUB_NORM, "blk.%d.attn_sub_norm" },
+ { LLM_TENSOR_FFN_SUB_NORM, "blk.%d.ffn_sub_norm" },
+ { LLM_TENSOR_DEC_OUTPUT_NORM, "dec.output_norm" },
+ { LLM_TENSOR_DEC_ATTN_NORM, "dec.blk.%d.attn_norm" },
+ { LLM_TENSOR_DEC_ATTN_Q, "dec.blk.%d.attn_q" },
+ { LLM_TENSOR_DEC_ATTN_K, "dec.blk.%d.attn_k" },
+ { LLM_TENSOR_DEC_ATTN_V, "dec.blk.%d.attn_v" },
+ { LLM_TENSOR_DEC_ATTN_OUT, "dec.blk.%d.attn_o" },
+ { LLM_TENSOR_DEC_ATTN_REL_B, "dec.blk.%d.attn_rel_b" },
+ { LLM_TENSOR_DEC_CROSS_ATTN_NORM, "dec.blk.%d.cross_attn_norm" },
+ { LLM_TENSOR_DEC_CROSS_ATTN_Q, "dec.blk.%d.cross_attn_q" },
+ { LLM_TENSOR_DEC_CROSS_ATTN_K, "dec.blk.%d.cross_attn_k" },
+ { LLM_TENSOR_DEC_CROSS_ATTN_V, "dec.blk.%d.cross_attn_v" },
+ { LLM_TENSOR_DEC_CROSS_ATTN_OUT, "dec.blk.%d.cross_attn_o" },
+ { LLM_TENSOR_DEC_CROSS_ATTN_REL_B, "dec.blk.%d.cross_attn_rel_b" },
+ { LLM_TENSOR_DEC_FFN_NORM, "dec.blk.%d.ffn_norm" },
+ { LLM_TENSOR_DEC_FFN_GATE, "dec.blk.%d.ffn_gate" },
+ { LLM_TENSOR_DEC_FFN_DOWN, "dec.blk.%d.ffn_down" },
+ { LLM_TENSOR_DEC_FFN_UP, "dec.blk.%d.ffn_up" },
+ { LLM_TENSOR_ENC_ATTN_NORM, "enc.blk.%d.attn_norm" },
+ { LLM_TENSOR_ENC_ATTN_Q, "enc.blk.%d.attn_q" },
+ { LLM_TENSOR_ENC_ATTN_K, "enc.blk.%d.attn_k" },
+ { LLM_TENSOR_ENC_ATTN_V, "enc.blk.%d.attn_v" },
+ { LLM_TENSOR_ENC_ATTN_OUT, "enc.blk.%d.attn_o" },
+ { LLM_TENSOR_ENC_ATTN_REL_B, "enc.blk.%d.attn_rel_b" },
+ { LLM_TENSOR_ENC_FFN_NORM, "enc.blk.%d.ffn_norm" },
+ { LLM_TENSOR_ENC_FFN_GATE, "enc.blk.%d.ffn_gate" },
+ { LLM_TENSOR_ENC_FFN_DOWN, "enc.blk.%d.ffn_down" },
+ { LLM_TENSOR_ENC_FFN_UP, "enc.blk.%d.ffn_up" },
+ { LLM_TENSOR_TIME_MIX_W1, "blk.%d.time_mix_w1" },
+ { LLM_TENSOR_TIME_MIX_W2, "blk.%d.time_mix_w2" },
+ { LLM_TENSOR_TIME_MIX_LERP_X, "blk.%d.time_mix_lerp_x" },
+ { LLM_TENSOR_TIME_MIX_LERP_W, "blk.%d.time_mix_lerp_w" },
+ { LLM_TENSOR_TIME_MIX_LERP_K, "blk.%d.time_mix_lerp_k" },
+ { LLM_TENSOR_TIME_MIX_LERP_V, "blk.%d.time_mix_lerp_v" },
+ { LLM_TENSOR_TIME_MIX_LERP_R, "blk.%d.time_mix_lerp_r" },
+ { LLM_TENSOR_TIME_MIX_LERP_G, "blk.%d.time_mix_lerp_g" },
+ { LLM_TENSOR_TIME_MIX_LERP_FUSED, "blk.%d.time_mix_lerp_fused" },
+ { LLM_TENSOR_TIME_MIX_FIRST, "blk.%d.time_mix_first" },
+ { LLM_TENSOR_TIME_MIX_DECAY, "blk.%d.time_mix_decay" },
+ { LLM_TENSOR_TIME_MIX_DECAY_W1, "blk.%d.time_mix_decay_w1" },
+ { LLM_TENSOR_TIME_MIX_DECAY_W2, "blk.%d.time_mix_decay_w2" },
+ { LLM_TENSOR_TIME_MIX_KEY, "blk.%d.time_mix_key" },
+ { LLM_TENSOR_TIME_MIX_VALUE, "blk.%d.time_mix_value" },
+ { LLM_TENSOR_TIME_MIX_RECEPTANCE, "blk.%d.time_mix_receptance" },
+ { LLM_TENSOR_TIME_MIX_GATE, "blk.%d.time_mix_gate" },
+ { LLM_TENSOR_TIME_MIX_LN, "blk.%d.time_mix_ln" },
+ { LLM_TENSOR_TIME_MIX_OUTPUT, "blk.%d.time_mix_output" },
+ { LLM_TENSOR_CHANNEL_MIX_LERP_K, "blk.%d.channel_mix_lerp_k" },
+ { LLM_TENSOR_CHANNEL_MIX_LERP_R, "blk.%d.channel_mix_lerp_r" },
+ { LLM_TENSOR_CHANNEL_MIX_KEY, "blk.%d.channel_mix_key" },
+ { LLM_TENSOR_CHANNEL_MIX_VALUE, "blk.%d.channel_mix_value" },
+ { LLM_TENSOR_CHANNEL_MIX_RECEPTANCE, "blk.%d.channel_mix_receptance" },
+ { LLM_TENSOR_TIME_MIX_W0, "blk.%d.time_mix_w0" },
+ { LLM_TENSOR_TIME_MIX_A0, "blk.%d.time_mix_a0" },
+ { LLM_TENSOR_TIME_MIX_A1, "blk.%d.time_mix_a1" },
+ { LLM_TENSOR_TIME_MIX_A2, "blk.%d.time_mix_a2" },
+ { LLM_TENSOR_TIME_MIX_V0, "blk.%d.time_mix_v0" },
+ { LLM_TENSOR_TIME_MIX_V1, "blk.%d.time_mix_v1" },
+ { LLM_TENSOR_TIME_MIX_V2, "blk.%d.time_mix_v2" },
+ { LLM_TENSOR_TIME_MIX_G1, "blk.%d.time_mix_g1" },
+ { LLM_TENSOR_TIME_MIX_G2, "blk.%d.time_mix_g2" },
+ { LLM_TENSOR_TIME_MIX_K_K, "blk.%d.time_mix_k_k" },
+ { LLM_TENSOR_TIME_MIX_K_A, "blk.%d.time_mix_k_a" },
+ { LLM_TENSOR_TIME_MIX_R_K, "blk.%d.time_mix_r_k" },
+ { LLM_TENSOR_CONV1D, "conv1d" },
+ { LLM_TENSOR_CONVNEXT_DW, "convnext.%d.dw" },
+ { LLM_TENSOR_CONVNEXT_NORM, "convnext.%d.norm" },
+ { LLM_TENSOR_CONVNEXT_PW1, "convnext.%d.pw1" },
+ { LLM_TENSOR_CONVNEXT_PW2, "convnext.%d.pw2" },
+ { LLM_TENSOR_CONVNEXT_GAMMA, "convnext.%d.gamma" },
+ { LLM_TENSOR_POS_NET_CONV1, "posnet.%d.conv1" },
+ { LLM_TENSOR_POS_NET_CONV2, "posnet.%d.conv2" },
+ { LLM_TENSOR_POS_NET_NORM, "posnet.%d.norm" },
+ { LLM_TENSOR_POS_NET_NORM1, "posnet.%d.norm1" },
+ { LLM_TENSOR_POS_NET_NORM2, "posnet.%d.norm2" },
+ { LLM_TENSOR_POS_NET_ATTN_NORM, "posnet.%d.attn_norm" },
+ { LLM_TENSOR_POS_NET_ATTN_Q, "posnet.%d.attn_q" },
+ { LLM_TENSOR_POS_NET_ATTN_K, "posnet.%d.attn_k" },
+ { LLM_TENSOR_POS_NET_ATTN_V, "posnet.%d.attn_v" },
+ { LLM_TENSOR_POS_NET_ATTN_OUT, "posnet.%d.attn_output" },
+ { LLM_TENSOR_ATTN_SINKS, "blk.%d.attn_sinks" },
+ { LLM_TENSOR_SHORTCONV_CONV, "blk.%d.shortconv.conv" },
+ { LLM_TENSOR_SHORTCONV_INPROJ, "blk.%d.shortconv.in_proj" },
+ { LLM_TENSOR_SHORTCONV_OUTPROJ, "blk.%d.shortconv.out_proj" },
+ { LLM_TENSOR_FFN_GATE_CHEXPS, "blk.%d.ffn_gate_chexps" },
+ { LLM_TENSOR_FFN_DOWN_CHEXPS, "blk.%d.ffn_down_chexps" },
+ { LLM_TENSOR_FFN_UP_CHEXPS, "blk.%d.ffn_up_chexps" },
+ { LLM_TENSOR_VISEXP_ATTN_QKV, "blk.%d.vis_attn_qkv" },
+ { LLM_TENSOR_VISEXP_ATTN_OUT, "blk.%d.vis_attn_output" },
+ { LLM_TENSOR_VISEXP_FFN_GATE, "blk.%d.vis_gate" },
+ { LLM_TENSOR_VISEXP_FFN_DOWN, "blk.%d.vis_down" },
+ { LLM_TENSOR_VISEXP_FFN_UP, "blk.%d.vis_up" },
};
+static std::set<llm_tensor> llm_get_tensor_names(llm_arch arch) {
+ switch (arch) {
+ case LLM_ARCH_CLIP:
+ return {};
+ case LLM_ARCH_LLAMA:
+ case LLM_ARCH_DECI:
+ case LLM_ARCH_MISTRAL3:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ROPE_FREQS,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_Q,
+ LLM_TENSOR_ATTN_K,
+ LLM_TENSOR_ATTN_V,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_ATTN_ROT_EMBD,
+ LLM_TENSOR_FFN_GATE_INP,
+ LLM_TENSOR_FFN_NORM,
+ LLM_TENSOR_FFN_GATE,
+ LLM_TENSOR_FFN_DOWN,
+ LLM_TENSOR_FFN_UP,
+ LLM_TENSOR_FFN_GATE_EXP,
+ LLM_TENSOR_FFN_DOWN_EXP,
+ LLM_TENSOR_FFN_UP_EXP,
+ LLM_TENSOR_FFN_GATE_EXPS,
+ LLM_TENSOR_FFN_DOWN_EXPS,
+ LLM_TENSOR_FFN_UP_EXPS,
+ };
+ case LLM_ARCH_ARCEE:
+ case LLM_ARCH_STARCODER2:
+ case LLM_ARCH_NEMOTRON:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ROPE_FREQS,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_Q,
+ LLM_TENSOR_ATTN_K,
+ LLM_TENSOR_ATTN_V,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_ATTN_ROT_EMBD,
+ LLM_TENSOR_FFN_NORM,
+ LLM_TENSOR_FFN_DOWN,
+ LLM_TENSOR_FFN_UP,
+ };
+ case LLM_ARCH_AFMOE:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_POST_NORM,
+ LLM_TENSOR_ATTN_Q,
+ LLM_TENSOR_ATTN_K,
+ LLM_TENSOR_ATTN_V,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_ATTN_Q_NORM,
+ LLM_TENSOR_ATTN_K_NORM,
+ LLM_TENSOR_ATTN_GATE,
+ LLM_TENSOR_FFN_NORM,
+ LLM_TENSOR_FFN_POST_NORM,
+ LLM_TENSOR_FFN_GATE_INP,
+ LLM_TENSOR_FFN_GATE,
+ LLM_TENSOR_FFN_DOWN,
+ LLM_TENSOR_FFN_UP,
+ LLM_TENSOR_FFN_GATE_EXPS,
+ LLM_TENSOR_FFN_DOWN_EXPS,
+ LLM_TENSOR_FFN_UP_EXPS,
+ LLM_TENSOR_FFN_GATE_SHEXP,
+ LLM_TENSOR_FFN_UP_SHEXP,
+ LLM_TENSOR_FFN_DOWN_SHEXP,
+ LLM_TENSOR_FFN_EXP_PROBS_B,
+ };
+ case LLM_ARCH_LLAMA4:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ROPE_FREQS,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_Q,
+ LLM_TENSOR_ATTN_K,
+ LLM_TENSOR_ATTN_V,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_ATTN_ROT_EMBD,
+ LLM_TENSOR_FFN_GATE_INP,
+ LLM_TENSOR_FFN_NORM,
+ LLM_TENSOR_FFN_GATE,
+ LLM_TENSOR_FFN_DOWN,
+ LLM_TENSOR_FFN_UP,
+ LLM_TENSOR_FFN_GATE_EXP,
+ LLM_TENSOR_FFN_DOWN_EXP,
+ LLM_TENSOR_FFN_UP_EXP,
+ LLM_TENSOR_FFN_GATE_EXPS,
+ LLM_TENSOR_FFN_DOWN_EXPS,
+ LLM_TENSOR_FFN_UP_EXPS,
+ LLM_TENSOR_FFN_GATE_SHEXP,
+ LLM_TENSOR_FFN_DOWN_SHEXP,
+ LLM_TENSOR_FFN_UP_SHEXP,
+ };
+ case LLM_ARCH_BAICHUAN:
+ case LLM_ARCH_ORION:
+ case LLM_ARCH_XVERSE:
+ case LLM_ARCH_EXAONE:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ROPE_FREQS,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_Q,
+ LLM_TENSOR_ATTN_K,
+ LLM_TENSOR_ATTN_V,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_ATTN_ROT_EMBD,
+ LLM_TENSOR_FFN_NORM,
+ LLM_TENSOR_FFN_GATE,
+ LLM_TENSOR_FFN_DOWN,
+ LLM_TENSOR_FFN_UP,
+ };
+ case LLM_ARCH_FALCON:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_NORM_2,
+ LLM_TENSOR_ATTN_QKV,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_FFN_DOWN,
+ LLM_TENSOR_FFN_UP,
+ };
+ case LLM_ARCH_GROK:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ROPE_FREQS,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_Q,
+ LLM_TENSOR_ATTN_K,
+ LLM_TENSOR_ATTN_V,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_ATTN_ROT_EMBD,
+ LLM_TENSOR_FFN_GATE_INP,
+ LLM_TENSOR_FFN_NORM,
+ LLM_TENSOR_FFN_GATE,
+ LLM_TENSOR_FFN_DOWN,
+ LLM_TENSOR_FFN_UP,
+ LLM_TENSOR_FFN_GATE_EXP,
+ LLM_TENSOR_FFN_DOWN_EXP,
+ LLM_TENSOR_FFN_UP_EXP,
+ LLM_TENSOR_FFN_GATE_EXPS,
+ LLM_TENSOR_FFN_DOWN_EXPS,
+ LLM_TENSOR_FFN_UP_EXPS,
+ LLM_TENSOR_FFN_POST_NORM,
+ LLM_TENSOR_LAYER_OUT_NORM,
+ LLM_TENSOR_ATTN_OUT_NORM,
+ };
+ case LLM_ARCH_GPT2:
+ case LLM_ARCH_STARCODER:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_POS_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_QKV,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_FFN_NORM,
+ LLM_TENSOR_FFN_UP,
+ LLM_TENSOR_FFN_DOWN,
+ };
+ case LLM_ARCH_GPTNEOX:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_QKV,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_FFN_NORM,
+ LLM_TENSOR_FFN_DOWN,
+ LLM_TENSOR_FFN_UP,
+ };
+ case LLM_ARCH_MPT:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_FFN_NORM,
+ LLM_TENSOR_ATTN_QKV,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_FFN_DOWN,
+ LLM_TENSOR_FFN_UP,
+ LLM_TENSOR_FFN_ACT,
+ LLM_TENSOR_POS_EMBD,
+ LLM_TENSOR_ATTN_Q_NORM,
+ LLM_TENSOR_ATTN_K_NORM,
+ };
+ case LLM_ARCH_REFACT:
+ case LLM_ARCH_QWEN2:
+ case LLM_ARCH_QWEN2VL:
+ case LLM_ARCH_INTERNLM2:
+ case LLM_ARCH_GRANITE:
+ case LLM_ARCH_ERNIE4_5:
+ case LLM_ARCH_SMOLLM3:
+ case LLM_ARCH_DREAM:
+ case LLM_ARCH_LLADA:
+ case LLM_ARCH_PANGU_EMBED:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_Q,
+ LLM_TENSOR_ATTN_K,
+ LLM_TENSOR_ATTN_V,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_FFN_NORM,
+ LLM_TENSOR_FFN_GATE,
+ LLM_TENSOR_FFN_DOWN,
+ LLM_TENSOR_FFN_UP,
+ };
+ case LLM_ARCH_BERT:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_TOKEN_EMBD_NORM,
+ LLM_TENSOR_TOKEN_TYPES,
+ LLM_TENSOR_POS_EMBD,
+ LLM_TENSOR_ATTN_OUT_NORM,
+ LLM_TENSOR_ATTN_QKV,
+ LLM_TENSOR_ATTN_Q,
+ LLM_TENSOR_ATTN_K,
+ LLM_TENSOR_ATTN_V,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_LAYER_OUT_NORM,
+ LLM_TENSOR_FFN_DOWN,
+ LLM_TENSOR_FFN_UP,
+ LLM_TENSOR_CLS,
+ LLM_TENSOR_CLS_OUT,
+ };
+ case LLM_ARCH_NOMIC_BERT:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_TOKEN_EMBD_NORM,
+ LLM_TENSOR_TOKEN_TYPES,
+ LLM_TENSOR_ATTN_OUT_NORM,
+ LLM_TENSOR_ATTN_QKV,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_LAYER_OUT_NORM,
+ LLM_TENSOR_FFN_GATE,
+ LLM_TENSOR_FFN_DOWN,
+ LLM_TENSOR_FFN_UP,
+ };
+ case LLM_ARCH_NOMIC_BERT_MOE:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_TOKEN_EMBD_NORM,
+ LLM_TENSOR_TOKEN_TYPES,
+ LLM_TENSOR_ATTN_OUT_NORM,
+ LLM_TENSOR_ATTN_QKV,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_LAYER_OUT_NORM,
+ LLM_TENSOR_FFN_GATE,
+ LLM_TENSOR_FFN_DOWN,
+ LLM_TENSOR_FFN_UP,
+ LLM_TENSOR_FFN_GATE_INP,
+ LLM_TENSOR_FFN_DOWN_EXPS,
+ LLM_TENSOR_FFN_UP_EXPS,
+ };
+ case LLM_ARCH_NEO_BERT:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_QKV,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_FFN_NORM,
+ LLM_TENSOR_FFN_DOWN,
+ LLM_TENSOR_FFN_UP,
+ LLM_TENSOR_ENC_OUTPUT_NORM,
+ LLM_TENSOR_CLS,
+ LLM_TENSOR_CLS_OUT,
+ };
+ case LLM_ARCH_JINA_BERT_V2:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_TOKEN_EMBD_NORM,
+ LLM_TENSOR_TOKEN_TYPES,
+ LLM_TENSOR_ATTN_NORM_2,
+ LLM_TENSOR_ATTN_OUT_NORM,
+ LLM_TENSOR_ATTN_Q,
+ LLM_TENSOR_ATTN_Q_NORM,
+ LLM_TENSOR_ATTN_K,
+ LLM_TENSOR_ATTN_K_NORM,
+ LLM_TENSOR_ATTN_V,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_LAYER_OUT_NORM,
+ LLM_TENSOR_FFN_DOWN,
+ LLM_TENSOR_FFN_GATE,
+ LLM_TENSOR_FFN_UP,
+ LLM_TENSOR_CLS,
+ };
+ case LLM_ARCH_JINA_BERT_V3:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_TOKEN_EMBD_NORM,
+ LLM_TENSOR_TOKEN_TYPES,
+ LLM_TENSOR_ATTN_OUT_NORM,
+ LLM_TENSOR_ATTN_QKV,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_FFN_DOWN,
+ LLM_TENSOR_FFN_UP,
+ LLM_TENSOR_LAYER_OUT_NORM,
+ };
+ case LLM_ARCH_BLOOM:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_TOKEN_EMBD_NORM,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_QKV,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_FFN_NORM,
+ LLM_TENSOR_FFN_UP,
+ LLM_TENSOR_FFN_DOWN,
+ };
+ case LLM_ARCH_STABLELM:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ROPE_FREQS,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_Q,
+ LLM_TENSOR_ATTN_K,
+ LLM_TENSOR_ATTN_V,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_FFN_NORM,
+ LLM_TENSOR_FFN_GATE,
+ LLM_TENSOR_FFN_DOWN,
+ LLM_TENSOR_FFN_UP,
+ LLM_TENSOR_ATTN_Q_NORM,
+ LLM_TENSOR_ATTN_K_NORM,
+ };
+ case LLM_ARCH_QWEN:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ROPE_FREQS,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_QKV,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_FFN_NORM,
+ LLM_TENSOR_FFN_GATE,
+ LLM_TENSOR_FFN_DOWN,
+ LLM_TENSOR_FFN_UP,
+ };
+ case LLM_ARCH_QWEN2MOE:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_Q,
+ LLM_TENSOR_ATTN_K,
+ LLM_TENSOR_ATTN_V,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_FFN_NORM,
+ LLM_TENSOR_FFN_GATE_INP,
+ LLM_TENSOR_FFN_GATE_EXPS,
+ LLM_TENSOR_FFN_DOWN_EXPS,
+ LLM_TENSOR_FFN_UP_EXPS,
+ LLM_TENSOR_FFN_GATE_INP_SHEXP,
+ LLM_TENSOR_FFN_GATE_SHEXP,
+ LLM_TENSOR_FFN_DOWN_SHEXP,
+ LLM_TENSOR_FFN_UP_SHEXP,
+ };
+ case LLM_ARCH_QWEN3:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_CLS_OUT,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_Q,
+ LLM_TENSOR_ATTN_Q_NORM,
+ LLM_TENSOR_ATTN_K,
+ LLM_TENSOR_ATTN_K_NORM,
+ LLM_TENSOR_ATTN_V,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_FFN_NORM,
+ LLM_TENSOR_FFN_GATE,
+ LLM_TENSOR_FFN_DOWN,
+ LLM_TENSOR_FFN_UP,
+ };
+ case LLM_ARCH_QWEN3MOE:
+ case LLM_ARCH_QWEN3VLMOE:
+ case LLM_ARCH_OLMOE:
+ case LLM_ARCH_LLADA_MOE:
+ case LLM_ARCH_RND1:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_Q,
+ LLM_TENSOR_ATTN_Q_NORM,
+ LLM_TENSOR_ATTN_K,
+ LLM_TENSOR_ATTN_K_NORM,
+ LLM_TENSOR_ATTN_V,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_FFN_NORM,
+ LLM_TENSOR_FFN_GATE_INP,
+ LLM_TENSOR_FFN_GATE_EXPS,
+ LLM_TENSOR_FFN_DOWN_EXPS,
+ LLM_TENSOR_FFN_UP_EXPS,
+ };
+ case LLM_ARCH_QWEN3NEXT:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_POST_NORM,
+ LLM_TENSOR_ATTN_Q,
+ LLM_TENSOR_ATTN_Q_NORM,
+ LLM_TENSOR_ATTN_K,
+ LLM_TENSOR_ATTN_K_NORM,
+ LLM_TENSOR_ATTN_V,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_FFN_NORM,
+ LLM_TENSOR_FFN_GATE_INP,
+ LLM_TENSOR_FFN_GATE_EXPS,
+ LLM_TENSOR_FFN_DOWN_EXPS,
+ LLM_TENSOR_FFN_UP_EXPS,
+ LLM_TENSOR_FFN_GATE_INP_SHEXP,
+ LLM_TENSOR_FFN_GATE_SHEXP,
+ LLM_TENSOR_FFN_DOWN_SHEXP,
+ LLM_TENSOR_FFN_UP_SHEXP,
+ LLM_TENSOR_SSM_A_NOSCAN,
+ LLM_TENSOR_SSM_CONV1D,
+ LLM_TENSOR_SSM_DT,
+ LLM_TENSOR_SSM_BETA_ALPHA,
+ LLM_TENSOR_SSM_IN,
+ LLM_TENSOR_SSM_NORM,
+ LLM_TENSOR_SSM_OUT,
+ };
+ case LLM_ARCH_QWEN3VL:
+ case LLM_ARCH_CHAMELEON:
+ case LLM_ARCH_HUNYUAN_DENSE:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_Q,
+ LLM_TENSOR_ATTN_Q_NORM,
+ LLM_TENSOR_ATTN_K,
+ LLM_TENSOR_ATTN_K_NORM,
+ LLM_TENSOR_ATTN_V,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_FFN_NORM,
+ LLM_TENSOR_FFN_GATE,
+ LLM_TENSOR_FFN_DOWN,
+ LLM_TENSOR_FFN_UP,
+ };
+ case LLM_ARCH_PHI2:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_QKV,
+ LLM_TENSOR_ATTN_Q,
+ LLM_TENSOR_ATTN_K,
+ LLM_TENSOR_ATTN_V,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_FFN_DOWN,
+ LLM_TENSOR_FFN_UP,
+ };
+ case LLM_ARCH_PHI3:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ROPE_FACTORS_LONG,
+ LLM_TENSOR_ROPE_FACTORS_SHORT,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_QKV,
+ LLM_TENSOR_ATTN_Q,
+ LLM_TENSOR_ATTN_K,
+ LLM_TENSOR_ATTN_V,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_FFN_NORM,
+ LLM_TENSOR_FFN_DOWN,
+ LLM_TENSOR_FFN_UP,
+ };
+ case LLM_ARCH_PHIMOE:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ROPE_FACTORS_LONG,
+ LLM_TENSOR_ROPE_FACTORS_SHORT,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_QKV,
+ LLM_TENSOR_ATTN_Q,
+ LLM_TENSOR_ATTN_K,
+ LLM_TENSOR_ATTN_V,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_FFN_NORM,
+ LLM_TENSOR_FFN_GATE_INP,
+ LLM_TENSOR_FFN_GATE_EXPS,
+ LLM_TENSOR_FFN_DOWN_EXPS,
+ LLM_TENSOR_FFN_UP_EXPS,
+ };
+ case LLM_ARCH_PLAMO:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ROPE_FREQS,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_Q,
+ LLM_TENSOR_ATTN_K,
+ LLM_TENSOR_ATTN_V,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_ATTN_ROT_EMBD,
+ LLM_TENSOR_FFN_GATE,
+ LLM_TENSOR_FFN_DOWN,
+ LLM_TENSOR_FFN_UP,
+ };
+ case LLM_ARCH_PLAMO2:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ROPE_FREQS,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_QKV,
+ LLM_TENSOR_ATTN_Q_NORM,
+ LLM_TENSOR_ATTN_K_NORM,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_ATTN_ROT_EMBD,
+ LLM_TENSOR_FFN_NORM,
+ LLM_TENSOR_FFN_DOWN,
+ LLM_TENSOR_FFN_UP,
+ LLM_TENSOR_SSM_IN,
+ LLM_TENSOR_SSM_CONV1D,
+ LLM_TENSOR_SSM_X,
+ LLM_TENSOR_SSM_DT,
+ LLM_TENSOR_SSM_A,
+ LLM_TENSOR_SSM_D,
+ LLM_TENSOR_SSM_OUT,
+ LLM_TENSOR_SSM_DT_NORM,
+ LLM_TENSOR_SSM_B_NORM,
+ LLM_TENSOR_SSM_C_NORM,
+ LLM_TENSOR_ATTN_POST_NORM,
+ LLM_TENSOR_FFN_POST_NORM,
+ };
+ case LLM_ARCH_CODESHELL:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ROPE_FREQS,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_Q,
+ LLM_TENSOR_ATTN_K,
+ LLM_TENSOR_ATTN_V,
+ LLM_TENSOR_ATTN_QKV,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_ATTN_ROT_EMBD,
+ LLM_TENSOR_FFN_NORM,
+ LLM_TENSOR_FFN_GATE,
+ LLM_TENSOR_FFN_DOWN,
+ LLM_TENSOR_FFN_UP,
+ };
+ case LLM_ARCH_MINICPM:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ROPE_FREQS,
+ LLM_TENSOR_ROPE_FACTORS_LONG,
+ LLM_TENSOR_ROPE_FACTORS_SHORT,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_Q,
+ LLM_TENSOR_ATTN_K,
+ LLM_TENSOR_ATTN_V,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_ATTN_ROT_EMBD,
+ LLM_TENSOR_FFN_GATE_INP,
+ LLM_TENSOR_FFN_NORM,
+ LLM_TENSOR_FFN_GATE,
+ LLM_TENSOR_FFN_DOWN,
+ LLM_TENSOR_FFN_UP,
+ LLM_TENSOR_FFN_GATE_EXP,
+ LLM_TENSOR_FFN_DOWN_EXP,
+ LLM_TENSOR_FFN_UP_EXP,
+ };
+ case LLM_ARCH_MINICPM3:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ROPE_FACTORS_LONG,
+ LLM_TENSOR_ROPE_FACTORS_SHORT,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_Q_A_NORM,
+ LLM_TENSOR_ATTN_KV_A_NORM,
+ LLM_TENSOR_ATTN_Q,
+ LLM_TENSOR_ATTN_Q_A,
+ LLM_TENSOR_ATTN_Q_B,
+ LLM_TENSOR_ATTN_KV_A_MQA,
+ LLM_TENSOR_ATTN_KV_B,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_FFN_NORM,
+ LLM_TENSOR_FFN_GATE,
+ LLM_TENSOR_FFN_UP,
+ LLM_TENSOR_FFN_DOWN,
+ };
+ case LLM_ARCH_GEMMA:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_Q,
+ LLM_TENSOR_ATTN_K,
+ LLM_TENSOR_ATTN_V,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_FFN_NORM,
+ LLM_TENSOR_FFN_GATE,
+ LLM_TENSOR_FFN_DOWN,
+ LLM_TENSOR_FFN_UP,
+ };
+ case LLM_ARCH_GEMMA2:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_Q,
+ LLM_TENSOR_ATTN_K,
+ LLM_TENSOR_ATTN_V,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_ATTN_POST_NORM,
+ LLM_TENSOR_FFN_NORM,
+ LLM_TENSOR_FFN_GATE,
+ LLM_TENSOR_FFN_DOWN,
+ LLM_TENSOR_FFN_UP,
+ LLM_TENSOR_FFN_POST_NORM,
+ };
+ case LLM_ARCH_GEMMA3:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_Q,
+ LLM_TENSOR_ATTN_Q_NORM,
+ LLM_TENSOR_ATTN_K,
+ LLM_TENSOR_ATTN_K_NORM,
+ LLM_TENSOR_ATTN_V,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_ATTN_POST_NORM,
+ LLM_TENSOR_FFN_NORM,
+ LLM_TENSOR_FFN_GATE,
+ LLM_TENSOR_FFN_DOWN,
+ LLM_TENSOR_FFN_UP,
+ LLM_TENSOR_FFN_POST_NORM,
+ };
+ case LLM_ARCH_GEMMA3N:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_Q,
+ LLM_TENSOR_ATTN_Q_NORM,
+ LLM_TENSOR_ATTN_K,
+ LLM_TENSOR_ATTN_K_NORM,
+ LLM_TENSOR_ATTN_V,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_ATTN_POST_NORM,
+ LLM_TENSOR_FFN_NORM,
+ LLM_TENSOR_FFN_GATE,
+ LLM_TENSOR_FFN_DOWN,
+ LLM_TENSOR_FFN_UP,
+ LLM_TENSOR_FFN_POST_NORM,
+ LLM_TENSOR_PER_LAYER_TOKEN_EMBD,
+ LLM_TENSOR_PER_LAYER_MODEL_PROJ,
+ LLM_TENSOR_PER_LAYER_PROJ_NORM,
+ LLM_TENSOR_ALTUP_UNEMBD_PROJ,
+ LLM_TENSOR_ALTUP_PROJ,
+ LLM_TENSOR_PER_LAYER_INP_GATE,
+ LLM_TENSOR_PER_LAYER_PROJ,
+ LLM_TENSOR_PER_LAYER_POST_NORM,
+ LLM_TENSOR_ALTUP_CORRECT_COEF,
+ LLM_TENSOR_ALTUP_CORRECT_SCALE,
+ LLM_TENSOR_ALTUP_PREDICT_COEF,
+ LLM_TENSOR_ALTUP_ROUTER,
+ LLM_TENSOR_ALTUP_ROUTER_NORM,
+ LLM_TENSOR_LAUREL_L,
+ LLM_TENSOR_LAUREL_R,
+ LLM_TENSOR_LAUREL_POST_NORM,
+ };
+ case LLM_ARCH_GEMMA_EMBEDDING:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_DENSE_2_OUT,
+ LLM_TENSOR_DENSE_3_OUT,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_Q,
+ LLM_TENSOR_ATTN_Q_NORM,
+ LLM_TENSOR_ATTN_K,
+ LLM_TENSOR_ATTN_K_NORM,
+ LLM_TENSOR_ATTN_V,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_ATTN_POST_NORM,
+ LLM_TENSOR_FFN_NORM,
+ LLM_TENSOR_FFN_GATE,
+ LLM_TENSOR_FFN_DOWN,
+ LLM_TENSOR_FFN_UP,
+ LLM_TENSOR_FFN_POST_NORM,
+ };
+ case LLM_ARCH_MAMBA:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_SSM_IN,
+ LLM_TENSOR_SSM_CONV1D,
+ LLM_TENSOR_SSM_X,
+ LLM_TENSOR_SSM_DT,
+ LLM_TENSOR_SSM_A,
+ LLM_TENSOR_SSM_D,
+ LLM_TENSOR_SSM_OUT,
+ };
+ case LLM_ARCH_MAMBA2:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_SSM_IN,
+ LLM_TENSOR_SSM_CONV1D,
+ LLM_TENSOR_SSM_DT,
+ LLM_TENSOR_SSM_A,
+ LLM_TENSOR_SSM_D,
+ LLM_TENSOR_SSM_NORM,
+ LLM_TENSOR_SSM_OUT,
+ };
+ case LLM_ARCH_JAMBA:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_SSM_IN,
+ LLM_TENSOR_SSM_CONV1D,
+ LLM_TENSOR_SSM_X,
+ LLM_TENSOR_SSM_DT,
+ LLM_TENSOR_SSM_DT_NORM,
+ LLM_TENSOR_SSM_A,
+ LLM_TENSOR_SSM_B_NORM,
+ LLM_TENSOR_SSM_C_NORM,
+ LLM_TENSOR_SSM_D,
+ LLM_TENSOR_SSM_OUT,
+ LLM_TENSOR_ATTN_Q,
+ LLM_TENSOR_ATTN_K,
+ LLM_TENSOR_ATTN_V,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_FFN_GATE_INP,
+ LLM_TENSOR_FFN_NORM,
+ LLM_TENSOR_FFN_GATE,
+ LLM_TENSOR_FFN_DOWN,
+ LLM_TENSOR_FFN_UP,
+ LLM_TENSOR_FFN_GATE_EXPS,
+ LLM_TENSOR_FFN_DOWN_EXPS,
+ LLM_TENSOR_FFN_UP_EXPS,
+ };
+ case LLM_ARCH_FALCON_H1:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_Q,
+ LLM_TENSOR_ATTN_K,
+ LLM_TENSOR_ATTN_V,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_SSM_IN,
+ LLM_TENSOR_SSM_CONV1D,
+ LLM_TENSOR_SSM_DT,
+ LLM_TENSOR_SSM_A,
+ LLM_TENSOR_SSM_D,
+ LLM_TENSOR_SSM_NORM,
+ LLM_TENSOR_SSM_OUT,
+ LLM_TENSOR_FFN_NORM,
+ LLM_TENSOR_FFN_GATE,
+ LLM_TENSOR_FFN_DOWN,
+ LLM_TENSOR_FFN_UP,
+ };
+ case LLM_ARCH_COMMAND_R:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_Q,
+ LLM_TENSOR_ATTN_K,
+ LLM_TENSOR_ATTN_V,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_FFN_GATE,
+ LLM_TENSOR_FFN_DOWN,
+ LLM_TENSOR_FFN_UP,
+ LLM_TENSOR_ATTN_Q_NORM,
+ LLM_TENSOR_ATTN_K_NORM,
+ };
+ case LLM_ARCH_COHERE2:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_Q,
+ LLM_TENSOR_ATTN_K,
+ LLM_TENSOR_ATTN_V,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_FFN_GATE,
+ LLM_TENSOR_FFN_DOWN,
+ LLM_TENSOR_FFN_UP,
+ };
+ case LLM_ARCH_DBRX:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ATTN_QKV,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_ATTN_OUT_NORM,
+ LLM_TENSOR_FFN_GATE_INP,
+ LLM_TENSOR_FFN_GATE_EXPS,
+ LLM_TENSOR_FFN_DOWN_EXPS,
+ LLM_TENSOR_FFN_UP_EXPS,
+ };
+ case LLM_ARCH_OLMO:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ATTN_Q,
+ LLM_TENSOR_ATTN_K,
+ LLM_TENSOR_ATTN_V,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_FFN_GATE,
+ LLM_TENSOR_FFN_DOWN,
+ LLM_TENSOR_FFN_UP,
+ };
+ case LLM_ARCH_OLMO2:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ATTN_Q,
+ LLM_TENSOR_ATTN_K,
+ LLM_TENSOR_ATTN_V,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_ATTN_POST_NORM,
+ LLM_TENSOR_ATTN_Q_NORM,
+ LLM_TENSOR_ATTN_K_NORM,
+ LLM_TENSOR_FFN_POST_NORM,
+ LLM_TENSOR_FFN_GATE,
+ LLM_TENSOR_FFN_DOWN,
+ LLM_TENSOR_FFN_UP,
+ };
+ case LLM_ARCH_OPENELM:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_QKV,
+ LLM_TENSOR_ATTN_Q_NORM,
+ LLM_TENSOR_ATTN_K_NORM,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_FFN_NORM,
+ LLM_TENSOR_FFN_GATE,
+ LLM_TENSOR_FFN_DOWN,
+ LLM_TENSOR_FFN_UP,
+ };
+ case LLM_ARCH_ARCTIC:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_Q,
+ LLM_TENSOR_ATTN_K,
+ LLM_TENSOR_ATTN_V,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_FFN_GATE_INP,
+ LLM_TENSOR_FFN_NORM,
+ LLM_TENSOR_FFN_GATE,
+ LLM_TENSOR_FFN_DOWN,
+ LLM_TENSOR_FFN_UP,
+ LLM_TENSOR_FFN_NORM_EXPS,
+ LLM_TENSOR_FFN_GATE_EXPS,
+ LLM_TENSOR_FFN_DOWN_EXPS,
+ LLM_TENSOR_FFN_UP_EXPS,
+ };
+ case LLM_ARCH_DEEPSEEK:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ROPE_FREQS,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_Q,
+ LLM_TENSOR_ATTN_K,
+ LLM_TENSOR_ATTN_V,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_ATTN_ROT_EMBD,
+ LLM_TENSOR_FFN_GATE_INP,
+ LLM_TENSOR_FFN_NORM,
+ LLM_TENSOR_FFN_GATE,
+ LLM_TENSOR_FFN_DOWN,
+ LLM_TENSOR_FFN_UP,
+ LLM_TENSOR_FFN_GATE_EXPS,
+ LLM_TENSOR_FFN_DOWN_EXPS,
+ LLM_TENSOR_FFN_UP_EXPS,
+ LLM_TENSOR_FFN_GATE_INP_SHEXP,
+ LLM_TENSOR_FFN_GATE_SHEXP,
+ LLM_TENSOR_FFN_DOWN_SHEXP,
+ LLM_TENSOR_FFN_UP_SHEXP,
+ };
+ case LLM_ARCH_DEEPSEEK2:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_Q_A_NORM,
+ LLM_TENSOR_ATTN_KV_A_NORM,
+ LLM_TENSOR_ATTN_Q,
+ LLM_TENSOR_ATTN_Q_A,
+ LLM_TENSOR_ATTN_Q_B,
+ LLM_TENSOR_ATTN_KV_A_MQA,
+ LLM_TENSOR_ATTN_KV_B,
+ LLM_TENSOR_ATTN_K_B,
+ LLM_TENSOR_ATTN_V_B,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_FFN_NORM,
+ LLM_TENSOR_FFN_GATE,
+ LLM_TENSOR_FFN_UP,
+ LLM_TENSOR_FFN_DOWN,
+ LLM_TENSOR_FFN_GATE_INP,
+ LLM_TENSOR_FFN_GATE_EXPS,
+ LLM_TENSOR_FFN_DOWN_EXPS,
+ LLM_TENSOR_FFN_UP_EXPS,
+ LLM_TENSOR_FFN_GATE_INP_SHEXP,
+ LLM_TENSOR_FFN_GATE_SHEXP,
+ LLM_TENSOR_FFN_DOWN_SHEXP,
+ LLM_TENSOR_FFN_UP_SHEXP,
+ LLM_TENSOR_FFN_EXP_PROBS_B,
+ };
+ case LLM_ARCH_PLM:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_Q,
+ LLM_TENSOR_ATTN_KV_A_MQA,
+ LLM_TENSOR_ATTN_KV_A_NORM,
+ LLM_TENSOR_ATTN_KV_B,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_FFN_NORM,
+ LLM_TENSOR_FFN_DOWN,
+ LLM_TENSOR_FFN_UP,
+ };
+ case LLM_ARCH_CHATGLM:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_ROPE_FREQS,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_QKV,
+ LLM_TENSOR_ATTN_Q,
+ LLM_TENSOR_ATTN_K,
+ LLM_TENSOR_ATTN_V,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_FFN_NORM,
+ LLM_TENSOR_FFN_UP,
+ LLM_TENSOR_FFN_DOWN,
+ };
+ case LLM_ARCH_GLM4:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_ROPE_FREQS,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_Q,
+ LLM_TENSOR_ATTN_K,
+ LLM_TENSOR_ATTN_V,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_FFN_NORM,
+ LLM_TENSOR_FFN_UP,
+ LLM_TENSOR_FFN_DOWN,
+ LLM_TENSOR_ATTN_POST_NORM,
+ LLM_TENSOR_FFN_POST_NORM,
+ };
+ case LLM_ARCH_GLM4_MOE:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_POST_NORM,
+ LLM_TENSOR_ATTN_Q,
+ LLM_TENSOR_ATTN_K,
+ LLM_TENSOR_ATTN_V,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_ATTN_Q_NORM,
+ LLM_TENSOR_ATTN_K_NORM,
+ LLM_TENSOR_FFN_GATE,
+ LLM_TENSOR_FFN_DOWN,
+ LLM_TENSOR_FFN_UP,
+ LLM_TENSOR_FFN_GATE_INP,
+ LLM_TENSOR_FFN_GATE_EXPS,
+ LLM_TENSOR_FFN_DOWN_EXPS,
+ LLM_TENSOR_FFN_UP_EXPS,
+ LLM_TENSOR_FFN_GATE_SHEXP,
+ LLM_TENSOR_FFN_DOWN_SHEXP,
+ LLM_TENSOR_FFN_UP_SHEXP,
+ LLM_TENSOR_FFN_EXP_PROBS_B,
+ LLM_TENSOR_NEXTN_EH_PROJ,
+ LLM_TENSOR_NEXTN_EMBED_TOKENS,
+ LLM_TENSOR_NEXTN_ENORM,
+ LLM_TENSOR_NEXTN_HNORM,
+ LLM_TENSOR_NEXTN_SHARED_HEAD_HEAD,
+ LLM_TENSOR_NEXTN_SHARED_HEAD_NORM,
+ };
+ case LLM_ARCH_BITNET:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_ATTN_Q,
+ LLM_TENSOR_ATTN_K,
+ LLM_TENSOR_ATTN_V,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_SUB_NORM,
+ LLM_TENSOR_FFN_GATE,
+ LLM_TENSOR_FFN_DOWN,
+ LLM_TENSOR_FFN_UP,
+ LLM_TENSOR_FFN_NORM,
+ LLM_TENSOR_FFN_SUB_NORM,
+ };
+ case LLM_ARCH_T5:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_DEC_OUTPUT_NORM,
+ LLM_TENSOR_DEC_ATTN_NORM,
+ LLM_TENSOR_DEC_ATTN_Q,
+ LLM_TENSOR_DEC_ATTN_K,
+ LLM_TENSOR_DEC_ATTN_V,
+ LLM_TENSOR_DEC_ATTN_OUT,
+ LLM_TENSOR_DEC_ATTN_REL_B,
+ LLM_TENSOR_DEC_CROSS_ATTN_NORM,
+ LLM_TENSOR_DEC_CROSS_ATTN_Q,
+ LLM_TENSOR_DEC_CROSS_ATTN_K,
+ LLM_TENSOR_DEC_CROSS_ATTN_V,
+ LLM_TENSOR_DEC_CROSS_ATTN_OUT,
+ LLM_TENSOR_DEC_CROSS_ATTN_REL_B,
+ LLM_TENSOR_DEC_FFN_NORM,
+ LLM_TENSOR_DEC_FFN_GATE,
+ LLM_TENSOR_DEC_FFN_DOWN,
+ LLM_TENSOR_DEC_FFN_UP,
+ LLM_TENSOR_ENC_OUTPUT_NORM,
+ LLM_TENSOR_ENC_ATTN_NORM,
+ LLM_TENSOR_ENC_ATTN_Q,
+ LLM_TENSOR_ENC_ATTN_K,
+ LLM_TENSOR_ENC_ATTN_V,
+ LLM_TENSOR_ENC_ATTN_OUT,
+ LLM_TENSOR_ENC_ATTN_REL_B,
+ LLM_TENSOR_ENC_FFN_NORM,
+ LLM_TENSOR_ENC_FFN_GATE,
+ LLM_TENSOR_ENC_FFN_DOWN,
+ LLM_TENSOR_ENC_FFN_UP,
+ };
+ case LLM_ARCH_T5ENCODER:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ENC_OUTPUT_NORM,
+ LLM_TENSOR_ENC_ATTN_NORM,
+ LLM_TENSOR_ENC_ATTN_Q,
+ LLM_TENSOR_ENC_ATTN_K,
+ LLM_TENSOR_ENC_ATTN_V,
+ LLM_TENSOR_ENC_ATTN_OUT,
+ LLM_TENSOR_ENC_ATTN_REL_B,
+ LLM_TENSOR_ENC_FFN_NORM,
+ LLM_TENSOR_ENC_FFN_GATE,
+ LLM_TENSOR_ENC_FFN_DOWN,
+ LLM_TENSOR_ENC_FFN_UP,
+ };
+ case LLM_ARCH_JAIS:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_QKV,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_FFN_NORM,
+ LLM_TENSOR_FFN_UP,
+ LLM_TENSOR_FFN_GATE,
+ LLM_TENSOR_FFN_DOWN,
+ };
+ case LLM_ARCH_NEMOTRON_H:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_SSM_IN,
+ LLM_TENSOR_SSM_CONV1D,
+ LLM_TENSOR_SSM_DT,
+ LLM_TENSOR_SSM_A,
+ LLM_TENSOR_SSM_D,
+ LLM_TENSOR_SSM_NORM,
+ LLM_TENSOR_SSM_OUT,
+ LLM_TENSOR_ATTN_Q,
+ LLM_TENSOR_ATTN_K,
+ LLM_TENSOR_ATTN_V,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_FFN_DOWN,
+ LLM_TENSOR_FFN_UP,
+ };
+ case LLM_ARCH_NEMOTRON_H_MOE:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ATTN_NORM,
+ // mamba(2) ssm layers
+ LLM_TENSOR_SSM_IN,
+ LLM_TENSOR_SSM_CONV1D,
+ LLM_TENSOR_SSM_DT,
+ LLM_TENSOR_SSM_A,
+ LLM_TENSOR_SSM_D,
+ LLM_TENSOR_SSM_NORM,
+ LLM_TENSOR_SSM_OUT,
+ // attention layers
+ LLM_TENSOR_ATTN_Q,
+ LLM_TENSOR_ATTN_K,
+ LLM_TENSOR_ATTN_V,
+ LLM_TENSOR_ATTN_OUT,
+ // dense FFN
+ LLM_TENSOR_FFN_DOWN,
+ LLM_TENSOR_FFN_UP,
+ // MoE FFN (for MoE layers)
+ LLM_TENSOR_FFN_GATE_INP,
+ LLM_TENSOR_FFN_UP_EXPS,
+ LLM_TENSOR_FFN_DOWN_EXPS,
+ LLM_TENSOR_FFN_EXP_PROBS_B,
+ // MoE shared expert layer
+ LLM_TENSOR_FFN_DOWN_SHEXP,
+ LLM_TENSOR_FFN_UP_SHEXP,
+ };
+ case LLM_ARCH_EXAONE4:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ROPE_FREQS,
+ LLM_TENSOR_ATTN_Q,
+ LLM_TENSOR_ATTN_Q_NORM,
+ LLM_TENSOR_ATTN_K,
+ LLM_TENSOR_ATTN_K_NORM,
+ LLM_TENSOR_ATTN_V,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_ATTN_POST_NORM,
+ LLM_TENSOR_FFN_GATE,
+ LLM_TENSOR_FFN_DOWN,
+ LLM_TENSOR_FFN_UP,
+ LLM_TENSOR_FFN_POST_NORM,
+ };
+ case LLM_ARCH_RWKV6:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_TOKEN_EMBD_NORM,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_NORM_2,
+ LLM_TENSOR_TIME_MIX_W1,
+ LLM_TENSOR_TIME_MIX_W2,
+ LLM_TENSOR_TIME_MIX_LERP_X,
+ LLM_TENSOR_TIME_MIX_LERP_W,
+ LLM_TENSOR_TIME_MIX_LERP_K,
+ LLM_TENSOR_TIME_MIX_LERP_V,
+ LLM_TENSOR_TIME_MIX_LERP_R,
+ LLM_TENSOR_TIME_MIX_LERP_G,
+ LLM_TENSOR_TIME_MIX_LERP_FUSED,
+ LLM_TENSOR_TIME_MIX_FIRST,
+ LLM_TENSOR_TIME_MIX_DECAY,
+ LLM_TENSOR_TIME_MIX_DECAY_W1,
+ LLM_TENSOR_TIME_MIX_DECAY_W2,
+ LLM_TENSOR_TIME_MIX_KEY,
+ LLM_TENSOR_TIME_MIX_VALUE,
+ LLM_TENSOR_TIME_MIX_RECEPTANCE,
+ LLM_TENSOR_TIME_MIX_GATE,
+ LLM_TENSOR_TIME_MIX_LN,
+ LLM_TENSOR_TIME_MIX_OUTPUT,
+ LLM_TENSOR_CHANNEL_MIX_LERP_K,
+ LLM_TENSOR_CHANNEL_MIX_LERP_R,
+ LLM_TENSOR_CHANNEL_MIX_KEY,
+ LLM_TENSOR_CHANNEL_MIX_VALUE,
+ LLM_TENSOR_CHANNEL_MIX_RECEPTANCE,
+ };
+ case LLM_ARCH_RWKV6QWEN2:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_TIME_MIX_W1,
+ LLM_TENSOR_TIME_MIX_W2,
+ LLM_TENSOR_TIME_MIX_LERP_X,
+ LLM_TENSOR_TIME_MIX_LERP_FUSED,
+ LLM_TENSOR_TIME_MIX_FIRST,
+ LLM_TENSOR_TIME_MIX_DECAY,
+ LLM_TENSOR_TIME_MIX_DECAY_W1,
+ LLM_TENSOR_TIME_MIX_DECAY_W2,
+ LLM_TENSOR_TIME_MIX_KEY,
+ LLM_TENSOR_TIME_MIX_VALUE,
+ LLM_TENSOR_TIME_MIX_RECEPTANCE,
+ LLM_TENSOR_TIME_MIX_GATE,
+ LLM_TENSOR_TIME_MIX_OUTPUT,
+ LLM_TENSOR_FFN_NORM,
+ LLM_TENSOR_FFN_GATE,
+ LLM_TENSOR_FFN_DOWN,
+ LLM_TENSOR_FFN_UP,
+ };
+ case LLM_ARCH_RWKV7:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_TOKEN_EMBD_NORM,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_NORM_2,
+ LLM_TENSOR_TIME_MIX_W0,
+ LLM_TENSOR_TIME_MIX_W1,
+ LLM_TENSOR_TIME_MIX_W2,
+ LLM_TENSOR_TIME_MIX_A0,
+ LLM_TENSOR_TIME_MIX_A1,
+ LLM_TENSOR_TIME_MIX_A2,
+ LLM_TENSOR_TIME_MIX_V0,
+ LLM_TENSOR_TIME_MIX_V1,
+ LLM_TENSOR_TIME_MIX_V2,
+ LLM_TENSOR_TIME_MIX_G1,
+ LLM_TENSOR_TIME_MIX_G2,
+ LLM_TENSOR_TIME_MIX_K_K,
+ LLM_TENSOR_TIME_MIX_K_A,
+ LLM_TENSOR_TIME_MIX_R_K,
+ LLM_TENSOR_TIME_MIX_LERP_FUSED,
+ LLM_TENSOR_TIME_MIX_KEY,
+ LLM_TENSOR_TIME_MIX_VALUE,
+ LLM_TENSOR_TIME_MIX_RECEPTANCE,
+ LLM_TENSOR_TIME_MIX_LN,
+ LLM_TENSOR_TIME_MIX_OUTPUT,
+ LLM_TENSOR_CHANNEL_MIX_LERP_K,
+ LLM_TENSOR_CHANNEL_MIX_KEY,
+ LLM_TENSOR_CHANNEL_MIX_VALUE,
+ };
+ case LLM_ARCH_ARWKV7:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_TOKEN_EMBD_NORM,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_TIME_MIX_W0,
+ LLM_TENSOR_TIME_MIX_W1,
+ LLM_TENSOR_TIME_MIX_W2,
+ LLM_TENSOR_TIME_MIX_A0,
+ LLM_TENSOR_TIME_MIX_A1,
+ LLM_TENSOR_TIME_MIX_A2,
+ LLM_TENSOR_TIME_MIX_V0,
+ LLM_TENSOR_TIME_MIX_V1,
+ LLM_TENSOR_TIME_MIX_V2,
+ LLM_TENSOR_TIME_MIX_G1,
+ LLM_TENSOR_TIME_MIX_G2,
+ LLM_TENSOR_TIME_MIX_K_K,
+ LLM_TENSOR_TIME_MIX_K_A,
+ LLM_TENSOR_TIME_MIX_R_K,
+ LLM_TENSOR_TIME_MIX_LERP_FUSED,
+ LLM_TENSOR_TIME_MIX_KEY,
+ LLM_TENSOR_TIME_MIX_VALUE,
+ LLM_TENSOR_TIME_MIX_RECEPTANCE,
+ LLM_TENSOR_TIME_MIX_LN,
+ LLM_TENSOR_TIME_MIX_OUTPUT,
+ LLM_TENSOR_FFN_NORM,
+ LLM_TENSOR_FFN_GATE,
+ LLM_TENSOR_FFN_DOWN,
+ LLM_TENSOR_FFN_UP,
+ };
+ case LLM_ARCH_GRANITE_MOE:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_Q,
+ LLM_TENSOR_ATTN_K,
+ LLM_TENSOR_ATTN_V,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_FFN_NORM,
+ LLM_TENSOR_FFN_GATE_INP,
+ LLM_TENSOR_FFN_GATE_EXPS,
+ LLM_TENSOR_FFN_DOWN_EXPS,
+ LLM_TENSOR_FFN_UP_EXPS,
+ LLM_TENSOR_FFN_GATE_SHEXP,
+ LLM_TENSOR_FFN_DOWN_SHEXP,
+ LLM_TENSOR_FFN_UP_SHEXP,
+ };
+ case LLM_ARCH_GRANITE_HYBRID:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_SSM_IN,
+ LLM_TENSOR_SSM_CONV1D,
+ LLM_TENSOR_SSM_DT,
+ LLM_TENSOR_SSM_A,
+ LLM_TENSOR_SSM_D,
+ LLM_TENSOR_SSM_NORM,
+ LLM_TENSOR_SSM_OUT,
+ LLM_TENSOR_ATTN_Q,
+ LLM_TENSOR_ATTN_K,
+ LLM_TENSOR_ATTN_V,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_FFN_NORM,
+ LLM_TENSOR_FFN_GATE,
+ LLM_TENSOR_FFN_DOWN,
+ LLM_TENSOR_FFN_UP,
+ LLM_TENSOR_FFN_NORM,
+ LLM_TENSOR_FFN_GATE_INP,
+ LLM_TENSOR_FFN_GATE_EXPS,
+ LLM_TENSOR_FFN_DOWN_EXPS,
+ LLM_TENSOR_FFN_UP_EXPS,
+ LLM_TENSOR_FFN_GATE_SHEXP,
+ LLM_TENSOR_FFN_DOWN_SHEXP,
+ LLM_TENSOR_FFN_UP_SHEXP,
+ };
+ case LLM_ARCH_WAVTOKENIZER_DEC:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_TOKEN_EMBD_NORM,
+ LLM_TENSOR_CONV1D,
+ LLM_TENSOR_CONVNEXT_DW,
+ LLM_TENSOR_CONVNEXT_NORM,
+ LLM_TENSOR_CONVNEXT_PW1,
+ LLM_TENSOR_CONVNEXT_PW2,
+ LLM_TENSOR_CONVNEXT_GAMMA,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_POS_NET_CONV1,
+ LLM_TENSOR_POS_NET_CONV2,
+ LLM_TENSOR_POS_NET_NORM,
+ LLM_TENSOR_POS_NET_NORM1,
+ LLM_TENSOR_POS_NET_NORM2,
+ LLM_TENSOR_POS_NET_ATTN_NORM,
+ LLM_TENSOR_POS_NET_ATTN_Q,
+ LLM_TENSOR_POS_NET_ATTN_K,
+ LLM_TENSOR_POS_NET_ATTN_V,
+ LLM_TENSOR_POS_NET_ATTN_OUT,
+ };
+ case LLM_ARCH_BAILINGMOE:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ROPE_FREQS,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_Q,
+ LLM_TENSOR_ATTN_K,
+ LLM_TENSOR_ATTN_V,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_FFN_GATE_INP,
+ LLM_TENSOR_FFN_NORM,
+ LLM_TENSOR_FFN_GATE_EXPS,
+ LLM_TENSOR_FFN_DOWN_EXPS,
+ LLM_TENSOR_FFN_UP_EXPS,
+ LLM_TENSOR_FFN_GATE_INP_SHEXP,
+ LLM_TENSOR_FFN_GATE_SHEXP,
+ LLM_TENSOR_FFN_DOWN_SHEXP,
+ LLM_TENSOR_FFN_UP_SHEXP,
+ };
+ case LLM_ARCH_BAILINGMOE2:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_Q_NORM,
+ LLM_TENSOR_ATTN_K_NORM,
+ LLM_TENSOR_ATTN_QKV,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_FFN_GATE_INP,
+ LLM_TENSOR_FFN_EXP_PROBS_B,
+ LLM_TENSOR_FFN_NORM,
+ LLM_TENSOR_FFN_GATE,
+ LLM_TENSOR_FFN_DOWN,
+ LLM_TENSOR_FFN_UP,
+ LLM_TENSOR_FFN_GATE_EXPS,
+ LLM_TENSOR_FFN_DOWN_EXPS,
+ LLM_TENSOR_FFN_UP_EXPS,
+ LLM_TENSOR_FFN_GATE_SHEXP,
+ LLM_TENSOR_FFN_DOWN_SHEXP,
+ LLM_TENSOR_FFN_UP_SHEXP,
+ LLM_TENSOR_NEXTN_EH_PROJ,
+ LLM_TENSOR_NEXTN_EMBED_TOKENS,
+ LLM_TENSOR_NEXTN_ENORM,
+ LLM_TENSOR_NEXTN_HNORM,
+ LLM_TENSOR_NEXTN_SHARED_HEAD_HEAD,
+ LLM_TENSOR_NEXTN_SHARED_HEAD_NORM,
+ LLM_TENSOR_LAYER_OUT_NORM,
+ };
+ case LLM_ARCH_DOTS1:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_Q,
+ LLM_TENSOR_ATTN_Q_NORM,
+ LLM_TENSOR_ATTN_K,
+ LLM_TENSOR_ATTN_K_NORM,
+ LLM_TENSOR_ATTN_V,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_FFN_NORM,
+ LLM_TENSOR_FFN_GATE,
+ LLM_TENSOR_FFN_UP,
+ LLM_TENSOR_FFN_DOWN,
+ LLM_TENSOR_FFN_GATE_INP,
+ LLM_TENSOR_FFN_GATE_EXPS,
+ LLM_TENSOR_FFN_DOWN_EXPS,
+ LLM_TENSOR_FFN_UP_EXPS,
+ LLM_TENSOR_FFN_GATE_INP_SHEXP,
+ LLM_TENSOR_FFN_GATE_SHEXP,
+ LLM_TENSOR_FFN_DOWN_SHEXP,
+ LLM_TENSOR_FFN_UP_SHEXP,
+ LLM_TENSOR_FFN_EXP_PROBS_B,
+ };
+ case LLM_ARCH_ERNIE4_5_MOE:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_Q,
+ LLM_TENSOR_ATTN_K,
+ LLM_TENSOR_ATTN_V,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_FFN_NORM,
+ LLM_TENSOR_FFN_GATE,
+ LLM_TENSOR_FFN_DOWN,
+ LLM_TENSOR_FFN_UP,
+ LLM_TENSOR_FFN_GATE_INP,
+ LLM_TENSOR_FFN_GATE_SHEXP,
+ LLM_TENSOR_FFN_DOWN_SHEXP,
+ LLM_TENSOR_FFN_UP_SHEXP,
+ LLM_TENSOR_FFN_GATE_EXPS,
+ LLM_TENSOR_FFN_DOWN_EXPS,
+ LLM_TENSOR_FFN_UP_EXPS,
+ LLM_TENSOR_FFN_EXP_PROBS_B,
+ };
+ case LLM_ARCH_HUNYUAN_MOE:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_Q,
+ LLM_TENSOR_ATTN_Q_NORM,
+ LLM_TENSOR_ATTN_K,
+ LLM_TENSOR_ATTN_K_NORM,
+ LLM_TENSOR_ATTN_V,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_FFN_GATE_INP,
+ LLM_TENSOR_FFN_NORM,
+ LLM_TENSOR_FFN_GATE_SHEXP,
+ LLM_TENSOR_FFN_DOWN_SHEXP,
+ LLM_TENSOR_FFN_UP_SHEXP,
+ LLM_TENSOR_FFN_GATE_EXPS,
+ LLM_TENSOR_FFN_DOWN_EXPS,
+ LLM_TENSOR_FFN_UP_EXPS,
+ };
+ case LLM_ARCH_OPENAI_MOE:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_POST_NORM,
+ LLM_TENSOR_ATTN_Q,
+ LLM_TENSOR_ATTN_K,
+ LLM_TENSOR_ATTN_V,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_ATTN_SINKS,
+ LLM_TENSOR_FFN_GATE_INP,
+ LLM_TENSOR_FFN_GATE_EXPS,
+ LLM_TENSOR_FFN_DOWN_EXPS,
+ LLM_TENSOR_FFN_UP_EXPS,
+ };
+ case LLM_ARCH_LFM2:
+ return {
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_Q,
+ LLM_TENSOR_ATTN_K,
+ LLM_TENSOR_ATTN_V,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_ATTN_K_NORM,
+ LLM_TENSOR_ATTN_Q_NORM,
+ LLM_TENSOR_FFN_DOWN,
+ LLM_TENSOR_FFN_GATE,
+ LLM_TENSOR_FFN_NORM,
+ LLM_TENSOR_FFN_UP,
+ LLM_TENSOR_SHORTCONV_CONV,
+ LLM_TENSOR_SHORTCONV_INPROJ,
+ LLM_TENSOR_SHORTCONV_OUTPROJ,
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM_LFM2,
+ LLM_TENSOR_OUTPUT,
+ };
+ case LLM_ARCH_LFM2MOE:
+ return {
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_Q,
+ LLM_TENSOR_ATTN_K,
+ LLM_TENSOR_ATTN_V,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_ATTN_K_NORM,
+ LLM_TENSOR_ATTN_Q_NORM,
+ LLM_TENSOR_FFN_DOWN,
+ LLM_TENSOR_FFN_GATE,
+ LLM_TENSOR_FFN_NORM,
+ LLM_TENSOR_FFN_UP,
+ LLM_TENSOR_SHORTCONV_CONV,
+ LLM_TENSOR_SHORTCONV_INPROJ,
+ LLM_TENSOR_SHORTCONV_OUTPROJ,
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM_LFM2,
+ LLM_TENSOR_FFN_GATE_INP,
+ LLM_TENSOR_FFN_GATE_EXPS,
+ LLM_TENSOR_FFN_DOWN_EXPS,
+ LLM_TENSOR_FFN_UP_EXPS,
+ LLM_TENSOR_FFN_EXP_PROBS_B,
+ };
+ case LLM_ARCH_SMALLTHINKER:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_Q,
+ LLM_TENSOR_ATTN_K,
+ LLM_TENSOR_ATTN_V,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_FFN_NORM,
+ LLM_TENSOR_FFN_GATE,
+ LLM_TENSOR_FFN_DOWN,
+ LLM_TENSOR_FFN_UP,
+ LLM_TENSOR_FFN_GATE_INP,
+ LLM_TENSOR_FFN_GATE_EXPS,
+ LLM_TENSOR_FFN_DOWN_EXPS,
+ LLM_TENSOR_FFN_UP_EXPS,
+ };
+ case LLM_ARCH_APERTUS:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ROPE_FREQS,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_Q,
+ LLM_TENSOR_ATTN_K,
+ LLM_TENSOR_ATTN_V,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_ATTN_Q_NORM,
+ LLM_TENSOR_ATTN_K_NORM,
+ LLM_TENSOR_FFN_NORM,
+ LLM_TENSOR_FFN_DOWN,
+ LLM_TENSOR_FFN_UP,
+ };
+ case LLM_ARCH_SEED_OSS:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_Q,
+ LLM_TENSOR_ATTN_K,
+ LLM_TENSOR_ATTN_V,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_ATTN_POST_NORM,
+ LLM_TENSOR_FFN_GATE,
+ LLM_TENSOR_FFN_DOWN,
+ LLM_TENSOR_FFN_UP,
+ };
+ case LLM_ARCH_GROVEMOE:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_Q,
+ LLM_TENSOR_ATTN_Q_NORM,
+ LLM_TENSOR_ATTN_K,
+ LLM_TENSOR_ATTN_K_NORM,
+ LLM_TENSOR_ATTN_V,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_FFN_NORM,
+ LLM_TENSOR_FFN_GATE_INP,
+ LLM_TENSOR_FFN_GATE_EXPS,
+ LLM_TENSOR_FFN_DOWN_EXPS,
+ LLM_TENSOR_FFN_UP_EXPS,
+ LLM_TENSOR_FFN_GATE_CHEXPS,
+ LLM_TENSOR_FFN_DOWN_CHEXPS,
+ LLM_TENSOR_FFN_UP_CHEXPS,
+ };
+ case LLM_ARCH_MINIMAX_M2:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_Q,
+ LLM_TENSOR_ATTN_K,
+ LLM_TENSOR_ATTN_V,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_ATTN_Q_NORM,
+ LLM_TENSOR_ATTN_K_NORM,
+ LLM_TENSOR_FFN_NORM,
+ LLM_TENSOR_FFN_GATE_INP,
+ LLM_TENSOR_FFN_GATE_EXPS,
+ LLM_TENSOR_FFN_DOWN_EXPS,
+ LLM_TENSOR_FFN_UP_EXPS,
+ LLM_TENSOR_FFN_EXP_PROBS_B,
+ };
+ case LLM_ARCH_COGVLM:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ LLM_TENSOR_OUTPUT_NORM,
+ LLM_TENSOR_OUTPUT,
+ LLM_TENSOR_ATTN_NORM,
+ LLM_TENSOR_ATTN_QKV,
+ LLM_TENSOR_ATTN_OUT,
+ LLM_TENSOR_FFN_NORM,
+ LLM_TENSOR_FFN_GATE,
+ LLM_TENSOR_FFN_DOWN,
+ LLM_TENSOR_FFN_UP,
+ LLM_TENSOR_VISEXP_ATTN_QKV,
+ LLM_TENSOR_VISEXP_ATTN_OUT,
+ LLM_TENSOR_VISEXP_FFN_GATE,
+ LLM_TENSOR_VISEXP_FFN_DOWN,
+ LLM_TENSOR_VISEXP_FFN_UP,
+ };
+ case LLM_ARCH_GPTJ:
+ case LLM_ARCH_UNKNOWN:
+ return {
+ LLM_TENSOR_TOKEN_EMBD,
+ };
+ default:
+ GGML_ABORT("unknown architecture for tensor mapping");
+ }
+}
+
// declare information about the model weight tensors:
// - the layer in which the tensor is going to be used. this is needed in order to assign the correct buffer type for the weight
// - the operator which is going to use the weight. this is needed to determine if the respective backend supports the operator
{LLM_TENSOR_DENSE_2_OUT, {LLM_TENSOR_LAYER_OUTPUT, GGML_OP_MUL_MAT}}, // Dense layer output
{LLM_TENSOR_DENSE_3_OUT, {LLM_TENSOR_LAYER_OUTPUT, GGML_OP_MUL_MAT}}, // Dense layer output
{LLM_TENSOR_OUTPUT_NORM, {LLM_TENSOR_LAYER_OUTPUT, GGML_OP_MUL}},
+ {LLM_TENSOR_OUTPUT_NORM_LFM2, {LLM_TENSOR_LAYER_OUTPUT, GGML_OP_MUL}},
{LLM_TENSOR_DEC_OUTPUT_NORM, {LLM_TENSOR_LAYER_OUTPUT, GGML_OP_MUL}},
{LLM_TENSOR_ENC_OUTPUT_NORM, {LLM_TENSOR_LAYER_OUTPUT, GGML_OP_MUL}},
{LLM_TENSOR_ROPE_FREQS, {LLM_TENSOR_LAYER_REPEATING, GGML_OP_ROPE}},
return name;
}
+LLM_TN_IMPL::LLM_TN_IMPL(llm_arch arch, llm_tensor tensor, const char * suffix, int bid, int xid)
+ : arch(arch), tensor(tensor), suffix(suffix), bid(bid), xid(xid),
+ model_tensors(llm_get_tensor_names(arch)) {}
+
std::string LLM_TN_IMPL::str() const {
- if (LLM_TENSOR_NAMES.at(arch).find(tensor) == LLM_TENSOR_NAMES.at(arch).end()) {
- return "__missing__";
+ if (LLM_TENSOR_NAMES.find(tensor) == LLM_TENSOR_NAMES.end()) {
+ GGML_ABORT("unknown tensor name for tensor id %d", static_cast<int>(tensor));
}
- std::string name = ::format(LLM_TENSOR_NAMES.at(arch).at(tensor), bid, xid);
+ if (model_tensors.find(tensor) == model_tensors.end()) {
+ return LLM_TENSOR_NAMES.at(tensor);
+ }
+ std::string name = ::format(LLM_TENSOR_NAMES.at(tensor), bid, xid);
if (suffix != nullptr) {
name += ".";
name += suffix;
case LLM_ARCH_LFM2:
case LLM_ARCH_LFM2MOE:
case LLM_ARCH_NEMOTRON_H:
+ case LLM_ARCH_NEMOTRON_H_MOE:
case LLM_ARCH_QWEN3NEXT:
return true;
default:
overview / pkg / ggml / sources / whisper.cpp / commitdiff