Update to Llama 2

fullstackbusiness · Jul 18, 2023 · cf29039 · cf29039
1 parent 80d28fb
commit cf29039
Show file tree

Hide file tree

Showing 3 changed files with 7 additions and 4 deletions.
diff --git a/config/config.yml b/config/config.yml
@@ -4,7 +4,9 @@ CHUNK_SIZE: 500
 CHUNK_OVERLAP: 50
 DATA_PATH: 'data/'
 DB_FAISS_PATH: 'vectorstore/db_faiss'
-MODEL_TYPE: 'mpt'
-MODEL_BIN_PATH: 'models/mpt-7b-instruct.ggmlv3.q8_0.bin'
+# MODEL_TYPE: 'mpt'
+# MODEL_BIN_PATH: 'models/mpt-7b-instruct.ggmlv3.q8_0.bin'
+MODEL_TYPE: 'llama'
+MODEL_BIN_PATH: 'models/llama-2-7b-chat.ggmlv3.q8_0.bin'
 MAX_NEW_TOKENS: 256
 TEMPERATURE: 0.01
diff --git a/models/model_download.txt b/models/model_download.txt
@@ -1 +1,2 @@
-Download the desired quantized MPT-7B-Instruct model from https://huggingface.co/TheBloke/MPT-7B-Instruct-GGML/tree/main
+Download the desired quantized MPT-7B-Instruct model from https://huggingface.co/TheBloke/MPT-7B-Instruct-GGML/tree/main
+Download the desired quantized Llama-2-7B-Chat model from https://huggingface.co/TheBloke/Llama-2-7B-Chat-GGML/tree/main
diff --git a/src/llm.py b/src/llm.py
@@ -17,7 +17,7 @@
 
 
 def build_llm():
-    # Local CTransformers MPT-7B-Instruct
+    # Local CTransformers model
     llm = CTransformers(model=cfg.MODEL_BIN_PATH,
                         model_type=cfg.MODEL_TYPE,
                         config={'max_new_tokens': cfg.MAX_NEW_TOKENS,