switch to smaller model to speed up CI test (#210)

Author: metascroy

.github/workflows/compile-gguf.yml

@@ -35,15 +35,16 @@ jobs:
       - name: Download GGUF
         run: |
           mkdir gguf_files
-          export GGUF_PATH=gguf_files/llama-2-7b.Q4_0.gguf
+          export GGUF_PATH=gguf_files/TinyLlama-1.1B-openorca.Q4_0.gguf
           export TOKENIZER_PATH=gguf_files/tokenizer.model
-          wget -O ${GGUF_PATH} "https://huggingface.co/TheBloke/Llama-2-7B-GGUF/resolve/main/llama-2-7b.Q4_0.gguf?download=true"
+
+          wget -O ${GGUF_PATH} "https://huggingface.co/TheBloke/TinyLlama-1.1B-1T-OpenOrca-GGUF/resolve/main/tinyllama-1.1b-1t-openorca.Q4_0.gguf?download=true"
           wget -O ${TOKENIZER_PATH} https://github.com/karpathy/llama2.c/raw/master/tokenizer.model
       - name: Run inference
         run: |
-          export GGUF_PATH=gguf_files/llama-2-7b.Q4_0.gguf
+          export GGUF_PATH=gguf_files/TinyLlama-1.1B-openorca.Q4_0.gguf
           export TOKENIZER_PATH=gguf_files/tokenizer.model
-          export MODEL_NAME=llama-2-7b.Q4_0.gguf
+          export MODEL_NAME=TinyLlama-1.1B-openorca.Q4_0.gguf
           export MODEL_DIR=/tmp
 
           echo "******************************************"