cyberes
/
local-llm-server
Archived
1
1
Fork 0
Code Issues Pull Requests Packages Projects Releases 1 Wiki Activity

Merge cluster to master #3

Merged
cyberes merged 163 commits from cluster into master 2023-10-27 19:19:22 -06:00
Conversation 0 Commits 163 Files Changed 91 +53 -53
1 changed files with 53 additions and 53 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files
Showing only changes of commit a37b12a221 - Show all commits

106
llm_server/routes/openai/chat_completions.py
View File

@ -57,7 +57,6 @@ def openai_chat_completions():
else:
handler.prompt = transform_messages_to_prompt(handler.request.json['messages'])
generated_text = ''
response_status_code = 0
start_time = time.time()
@ -98,61 +97,62 @@ def openai_chat_completions():
oai_string = generate_oai_string(30)
def generate():
response = generator(msg_to_backend, handler.backend_url)
generated_text = ''
partial_response = b''
for chunk in response.iter_content(chunk_size=1):
partial_response += chunk
if partial_response.endswith(b'\x00'):
json_strs = partial_response.split(b'\x00')
for json_str in json_strs:
if json_str:
try:
json_obj = json.loads(json_str.decode())
new = json_obj['text'][0].split(handler.prompt + generated_text)[1]
generated_text = generated_text + new
except IndexError:
# ????
continue
try:
response = generator(msg_to_backend, handler.backend_url)
generated_text = ''
partial_response = b''
for chunk in response.iter_content(chunk_size=1):
partial_response += chunk
if partial_response.endswith(b'\x00'):
json_strs = partial_response.split(b'\x00')
for json_str in json_strs:
if json_str:
try:
json_obj = json.loads(json_str.decode())
new = json_obj['text'][0].split(handler.prompt + generated_text)[1]
generated_text = generated_text + new
except IndexError:
# ????
continue
data = {
"id": f"chatcmpl-{oai_string}",
"object": "chat.completion.chunk",
"created": int(time.time()),
"model": model,
"choices": [
{
"index": 0,
"delta": {
"content": new
},
"finish_reason": None
}
]
}
yield f'data: {json.dumps(data)}\n\n'
yield 'data: [DONE]\n\n'
end_time = time.time()
elapsed_time = end_time - start_time
log_to_db(
handler.client_ip,
handler.token,
handler.prompt,
generated_text,
elapsed_time,
handler.parameters,
r_headers,
response_status_code,
r_url,
handler.backend_url,
)
data = {
"id": f"chatcmpl-{oai_string}",
"object": "chat.completion.chunk",
"created": int(time.time()),
"model": model,
"choices": [
{
"index": 0,
"delta": {
"content": new
},
"finish_reason": None
}
]
}
yield f'data: {json.dumps(data)}\n\n'
yield 'data: [DONE]\n\n'
end_time = time.time()
elapsed_time = end_time - start_time
log_to_db(
handler.client_ip,
handler.token,
handler.prompt,
generated_text,
elapsed_time,
handler.parameters,
r_headers,
response_status_code,
r_url,
handler.backend_url,
)
finally:
# The worker incremented it, we'll decrement it.
decrement_ip_count(handler.client_ip, 'processing_ips')
decr_active_workers(handler.selected_model, handler.backend_url)
print(len(generated_text))
return Response(generate(), mimetype='text/event-stream')
except Exception:
traceback.print_exc()
return 'INTERNAL SERVER', 500
finally:
# The worker incremented it, we'll decrement it.
decrement_ip_count(handler.client_ip, 'processing_ips')
decr_active_workers(handler.selected_model, handler.backend_url)
print(len(generated_text))