From: Tom Bailey Date: Sun, 23 Jul 2023 14:51:13 +0000 (+0900) Subject: readme : fix gpt-neox docs (#410) X-Git-Tag: upstream/0.0.1642~1302 X-Git-Url: https://git.djapps.eu/?a=commitdiff_plain;h=3accb95942d1f6b4f6b7b6a2e854b4f0beaef3d3;p=pkg%2Fggml%2Fsources%2Fggml readme : fix gpt-neox docs (#410) * Fix gpt neox example directory reference * Fix gpt neox bin directory references --- diff --git a/examples/gpt-neox/README.md b/examples/gpt-neox/README.md index f7a862ff..64c6d7c6 100644 --- a/examples/gpt-neox/README.md +++ b/examples/gpt-neox/README.md @@ -21,10 +21,10 @@ git clone https://huggingface.co/stabilityai/gpt_neox-base-alpha-3b python3 -m pip install -r ../requirements.txt # convert model to FP16 -python3 ../examples/gpt_neox/convert-h5-to-ggml.py ./stablelm-base-alpha-3b/ 1 +python3 ../examples/gpt-neox/convert-h5-to-ggml.py ./stablelm-base-alpha-3b/ 1 # run inference using FP16 precision -make -j && ./bin/gpt_neox -m ./stablelm-base-alpha-3b/ggml-model-f16.bin -p "I believe the meaning of life is" -t 8 -n 64 +make -j && ./bin/gpt-neox -m ./stablelm-base-alpha-3b/ggml-model-f16.bin -p "I believe the meaning of life is" -t 8 -n 64 main: seed = 1681940611 gpt_neox_model_load: loading model from 'models/stablelm-base-alpha-3b/ggml-model-f16.bin' - please wait ... @@ -63,10 +63,10 @@ main: total time = 6911.26 ms ```bash # quantize the model to 5-bits using Q5_0 quantization -./bin/gpt_neox-quantize ./stablelm-base-alpha-3b/ggml-model-f16.bin ./stablelm-base-alpha-3b/ggml-model-q5_0.bin q5_0 +./bin/gpt-neox-quantize ./stablelm-base-alpha-3b/ggml-model-f16.bin ./stablelm-base-alpha-3b/ggml-model-q5_0.bin q5_0 # run the quantized model -./bin/gpt_neox -m ./stablelm-base-alpha-3b/ggml-model-q5_0.bin -p "I believe the meaning of life is" -t 8 -n 64 +./bin/gpt-neox -m ./stablelm-base-alpha-3b/ggml-model-q5_0.bin -p "I believe the meaning of life is" -t 8 -n 64 main: seed = 1682021489 gpt_neox_model_load: loading model from 'models/stablelm-base-alpha-3b/ggml-model-q5_0.bin' - please wait ...