-
When using:
```
torchrun --nproc_per_node=2 --master_port=20001 fastchat/train/train.py \
--model_name_or_path lmsys/vicuna-7b-v1.5 \
--data_path data/dummy_conversation.json \
--bf…
-
### Describe your problem
"I have a very complex graph model, and in one of the processes, I connect an answer generation component with a human-computer interaction component. In this stage, I need …
SG-XM updated
3 months ago
-
If to inspect the JSON file with exported conversations, you could see the token cost for both inbound (prompts) and outbound (generated text) messages. But that information is not available on the h…
-
I successfully reproduce the notebook output for "mistralai/Mistral-7B-Instruct-v0.1".
But when I change the model, I cannot get desired result with the same setting.
Am I missing something? Or the …
-
The more I'm training the bot ... The worse are the responses....
Check it out : -
HUMAN ++++ Hi
BOT ++++ dennings dennings dennings dennings dennings shed grocer grocer grocer groce…
-
The youcook2 data repository (http://youcook2.eecs.umich.edu/download) only provides a script to download the raw videos into a folder `.../youcook2/raw_videos/`. However, the entries in the `youcook_…
-
```py
from unsloth import FastLanguageModel
from unsloth import is_bfloat16_supported
import torch
from unsloth.chat_templates import get_chat_template
from trl import SFTTrainer
from transform…
-
第一条数据就错了，只有直角三角形才能用勾股定理。这模型都被整数据的带沟里去了
moss/SFT_data/conversations/conversation_with_plugins/calculator/calculator_plugin_conversation_1.json
"chat": {
"turn_1": {
"H…
-
We have the PDF there and then they just have to press buttons on the pdf that will cycle through this pattern:
Pattern is
1. check clio first.
2. Guess the gaps using AI [from, EMAILS/ TEXT T…
-
***Placeholder feature issue -- requires further discussion***
This is a generalisation of several specific feature requests from CLI end users:
#89, #462, #695, #789, #837, #1001, #1043, #1697 an…