munish0838 commited on
Commit
d33b09e
โ€ข
1 Parent(s): 358425d

Create README.md

Browse files
Files changed (1) hide show
  1. README.md +72 -0
README.md ADDED
@@ -0,0 +1,72 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apache-2.0
3
+ base_model: TinyLlama/TinyLlama-1.1B-Chat-v1.0
4
+ datasets:
5
+ - cerebras/SlimPajama-627B
6
+ - bigcode/starcoderdata
7
+ - HuggingFaceH4/ultrachat_200k
8
+ - HuggingFaceH4/ultrafeedback_binarized
9
+ language:
10
+ - en
11
+ widget:
12
+ - example_title: Fibonacci (Python)
13
+ messages:
14
+ - role: system
15
+ content: You are a chatbot who can help code!
16
+ - role: user
17
+ content: >-
18
+ Write me a function to calculate the first 10 digits of the fibonacci
19
+ sequence in Python and print it out to the CLI.
20
+ pipeline_tag: text-generation
21
+ ---
22
+
23
+
24
+ # QuantFactory/TinyLlama-1.1B-Chat-v1.0-GGUF
25
+ This is quantized version of [TinyLlama/TinyLlama-1.1B-Chat-v1.0](https://huggingface.co/TinyLlama/TinyLlama-1.1B-Chat-v1.0) created using llama.cpp
26
+
27
+ # Model Description
28
+
29
+ https://github.com/jzhang38/TinyLlama
30
+
31
+ The TinyLlama project aims to **pretrain** a **1.1B Llama model on 3 trillion tokens**. With some proper optimization, we can achieve this within a span of "just" 90 days using 16 A100-40G GPUs ๐Ÿš€๐Ÿš€. The training has started on 2023-09-01.
32
+
33
+
34
+ We adopted exactly the same architecture and tokenizer as Llama 2. This means TinyLlama can be plugged and played in many open-source projects built upon Llama. Besides, TinyLlama is compact with only 1.1B parameters. This compactness allows it to cater to a multitude of applications demanding a restricted computation and memory footprint.
35
+
36
+ #### This Model
37
+ This is the chat model finetuned on top of [TinyLlama/TinyLlama-1.1B-intermediate-step-1431k-3T](https://huggingface.co/TinyLlama/TinyLlama-1.1B-intermediate-step-1431k-3T). **We follow [HF's Zephyr](https://huggingface.co/HuggingFaceH4/zephyr-7b-alpha)'s training recipe.** The model was " initially fine-tuned on a variant of the [`UltraChat`](https://huggingface.co/datasets/stingning/ultrachat) dataset, which contains a diverse range of synthetic dialogues generated by ChatGPT.
38
+ We then further aligned the model with [๐Ÿค— TRL's](https://github.com/huggingface/trl) `DPOTrainer` on the [openbmb/UltraFeedback](https://huggingface.co/datasets/openbmb/UltraFeedback) dataset, which contain 64k prompts and model completions that are ranked by GPT-4."
39
+
40
+
41
+ #### How to use
42
+ You will need the transformers>=4.34
43
+ Do check the [TinyLlama](https://github.com/jzhang38/TinyLlama) github page for more information.
44
+
45
+ ```python
46
+ # Install transformers from source - only needed for versions <= v4.34
47
+ # pip install git+https://github.com/huggingface/transformers.git
48
+ # pip install accelerate
49
+
50
+ import torch
51
+ from transformers import pipeline
52
+
53
+ pipe = pipeline("text-generation", model="TinyLlama/TinyLlama-1.1B-Chat-v1.0", torch_dtype=torch.bfloat16, device_map="auto")
54
+
55
+ # We use the tokenizer's chat template to format each message - see https://huggingface.co/docs/transformers/main/en/chat_templating
56
+ messages = [
57
+ {
58
+ "role": "system",
59
+ "content": "You are a friendly chatbot who always responds in the style of a pirate",
60
+ },
61
+ {"role": "user", "content": "How many helicopters can a human eat in one sitting?"},
62
+ ]
63
+ prompt = pipe.tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
64
+ outputs = pipe(prompt, max_new_tokens=256, do_sample=True, temperature=0.7, top_k=50, top_p=0.95)
65
+ print(outputs[0]["generated_text"])
66
+ # <|system|>
67
+ # You are a friendly chatbot who always responds in the style of a pirate.</s>
68
+ # <|user|>
69
+ # How many helicopters can a human eat in one sitting?</s>
70
+ # <|assistant|>
71
+ # ...
72
+ ```