Update app/llm.py
Browse files- app/llm.py +3 -3
app/llm.py
CHANGED
|
@@ -36,7 +36,7 @@ llm_chat = llama_cpp.Llama.from_pretrained(
|
|
| 36 |
verbose=False,
|
| 37 |
n_ctx=1024,
|
| 38 |
n_gpu_layers=0,
|
| 39 |
-
|
| 40 |
)
|
| 41 |
llm_generate = llama_cpp.Llama.from_pretrained(
|
| 42 |
repo_id="moriire/healthcare-GGUF",
|
|
@@ -47,7 +47,7 @@ llm_generate = llama_cpp.Llama.from_pretrained(
|
|
| 47 |
n_gpu_layers=0,
|
| 48 |
mirostat_mode=2,
|
| 49 |
mirostat_tau=4.0,
|
| 50 |
-
mirostat_eta=1.1
|
| 51 |
#chat_format="llama-2"
|
| 52 |
)
|
| 53 |
# Logger setup
|
|
@@ -87,7 +87,7 @@ async def chat(chatm:ChatModel):#, user: schemas.BaseUser = fastapi.Depends(curr
|
|
| 87 |
print(output)
|
| 88 |
#print(output)
|
| 89 |
et = time()
|
| 90 |
-
output["time"] = et - st
|
| 91 |
#messages.append({'role': "assistant", "content": output['choices'][0]['message']['content']})
|
| 92 |
#print(messages)
|
| 93 |
return output
|
|
|
|
| 36 |
verbose=False,
|
| 37 |
n_ctx=1024,
|
| 38 |
n_gpu_layers=0,
|
| 39 |
+
chat_format="llama-2"
|
| 40 |
)
|
| 41 |
llm_generate = llama_cpp.Llama.from_pretrained(
|
| 42 |
repo_id="moriire/healthcare-GGUF",
|
|
|
|
| 47 |
n_gpu_layers=0,
|
| 48 |
mirostat_mode=2,
|
| 49 |
mirostat_tau=4.0,
|
| 50 |
+
mirostat_eta=1.1,
|
| 51 |
#chat_format="llama-2"
|
| 52 |
)
|
| 53 |
# Logger setup
|
|
|
|
| 87 |
print(output)
|
| 88 |
#print(output)
|
| 89 |
et = time()
|
| 90 |
+
#output["time"] = et - st
|
| 91 |
#messages.append({'role': "assistant", "content": output['choices'][0]['message']['content']})
|
| 92 |
#print(messages)
|
| 93 |
return output
|