date3k2 commited on
Commit
df7dbb2
·
verified ·
1 Parent(s): d656362

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +35 -3
README.md CHANGED
@@ -2,6 +2,7 @@
2
  base_model: unsloth/llama-3-8b-Instruct-bnb-4bit
3
  language:
4
  - en
 
5
  license: apache-2.0
6
  tags:
7
  - text-generation-inference
@@ -10,14 +11,45 @@ tags:
10
  - llama
11
  - trl
12
  - sft
 
 
 
13
  ---
14
 
15
- # Uploaded model
16
 
17
  - **Developed by:** date3k2
18
  - **License:** apache-2.0
19
  - **Finetuned from model :** unsloth/llama-3-8b-Instruct-bnb-4bit
20
 
21
- This llama model was trained 2x faster with [Unsloth](https://github.com/unslothai/unsloth) and Huggingface's TRL library.
22
-
23
  [<img src="https://raw.githubusercontent.com/unslothai/unsloth/main/images/unsloth%20made%20with%20love.png" width="200"/>](https://github.com/unslothai/unsloth)
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
2
  base_model: unsloth/llama-3-8b-Instruct-bnb-4bit
3
  language:
4
  - en
5
+ - vi
6
  license: apache-2.0
7
  tags:
8
  - text-generation-inference
 
11
  - llama
12
  - trl
13
  - sft
14
+ datasets:
15
+ - bkai-foundation-models/vi-self-chat-sharegpt-format
16
+ pipeline_tag: text-generation
17
  ---
18
 
19
+ # Llama-8B-Instruct-ShareGPT
20
 
21
  - **Developed by:** date3k2
22
  - **License:** apache-2.0
23
  - **Finetuned from model :** unsloth/llama-3-8b-Instruct-bnb-4bit
24
 
 
 
25
  [<img src="https://raw.githubusercontent.com/unslothai/unsloth/main/images/unsloth%20made%20with%20love.png" width="200"/>](https://github.com/unslothai/unsloth)
26
+ [<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="200" height="32"/>](https://wandb.ai/date3k2/unsloth-llama-3-8b-instruct)
27
+ ## Usage
28
+
29
+ ```python
30
+ # !pip install "unsloth[colab-new] @ git+https://github.com/unslothai/unsloth.git"
31
+ # !pip install --no-deps xformers "trl<0.9.0" peft accelerate bitsandbytes
32
+ from unsloth import FastLanguageModel
33
+ model, tokenizer = FastLanguageModel.from_pretrained(
34
+ model_name = "date3k2/llama3-8b-instruct-sharegpt",
35
+ max_seq_length = 2048,
36
+ dtype = None,
37
+ load_in_4bit = True,
38
+ )
39
+ FastLanguageModel.for_inference(model) # Enable native 2x faster inference
40
+
41
+ messages = [
42
+ {"from": "human", "value": "Hãy gợi ý một số địa điểm du lịch ở Hà Nội."},
43
+ ]
44
+ inputs = tokenizer.apply_chat_template(
45
+ messages,
46
+ tokenize = True,
47
+ add_generation_prompt = True, # Must add for generation
48
+ return_tensors = "pt",
49
+ ).to("cuda")
50
+
51
+ from transformers import TextStreamer
52
+ text_streamer = TextStreamer(tokenizer)
53
+ _ = model.generate(input_ids = inputs, streamer = text_streamer, max_new_tokens = 512, use_cache = True)
54
+
55
+ ```