Hugging Face's logo

RichardErkhov
/
TinyLlama_-_TinyLlama-1.1B-Chat-v0.5-gguf

GGUF
Model card Files Community
RichardErkhov commited on
Commit
b1993cc
•
1 Parent(s): 7b12c99

uploaded readme

Browse files
Files changed (1) hide show
  1. README.md +106 -0
README.md ADDED
@@ -0,0 +1,106 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Quantization made by Richard Erkhov.
2
+
3
+ [Github](https://github.com/RichardErkhov)
4
+
5
+ [Discord](https://discord.gg/pvy7H8DZMG)
6
+
7
+ [Request more models](https://github.com/RichardErkhov/quant_request)
8
+
9
+
10
+ TinyLlama-1.1B-Chat-v0.5 - GGUF
11
+ - Model creator: https://huggingface.co/TinyLlama/
12
+ - Original model: https://huggingface.co/TinyLlama/TinyLlama-1.1B-Chat-v0.5/
13
+
14
+
15
+ | Name | Quant method | Size |
16
+ | ---- | ---- | ---- |
17
+ | [TinyLlama-1.1B-Chat-v0.5.Q2_K.gguf](https://huggingface.co/RichardErkhov/TinyLlama_-_TinyLlama-1.1B-Chat-v0.5-gguf/blob/main/TinyLlama-1.1B-Chat-v0.5.Q2_K.gguf) | Q2_K | 0.4GB |
18
+ | [TinyLlama-1.1B-Chat-v0.5.IQ3_XS.gguf](https://huggingface.co/RichardErkhov/TinyLlama_-_TinyLlama-1.1B-Chat-v0.5-gguf/blob/main/TinyLlama-1.1B-Chat-v0.5.IQ3_XS.gguf) | IQ3_XS | 0.44GB |
19
+ | [TinyLlama-1.1B-Chat-v0.5.IQ3_S.gguf](https://huggingface.co/RichardErkhov/TinyLlama_-_TinyLlama-1.1B-Chat-v0.5-gguf/blob/main/TinyLlama-1.1B-Chat-v0.5.IQ3_S.gguf) | IQ3_S | 0.47GB |
20
+ | [TinyLlama-1.1B-Chat-v0.5.Q3_K_S.gguf](https://huggingface.co/RichardErkhov/TinyLlama_-_TinyLlama-1.1B-Chat-v0.5-gguf/blob/main/TinyLlama-1.1B-Chat-v0.5.Q3_K_S.gguf) | Q3_K_S | 0.47GB |
21
+ | [TinyLlama-1.1B-Chat-v0.5.IQ3_M.gguf](https://huggingface.co/RichardErkhov/TinyLlama_-_TinyLlama-1.1B-Chat-v0.5-gguf/blob/main/TinyLlama-1.1B-Chat-v0.5.IQ3_M.gguf) | IQ3_M | 0.48GB |
22
+ | [TinyLlama-1.1B-Chat-v0.5.Q3_K.gguf](https://huggingface.co/RichardErkhov/TinyLlama_-_TinyLlama-1.1B-Chat-v0.5-gguf/blob/main/TinyLlama-1.1B-Chat-v0.5.Q3_K.gguf) | Q3_K | 0.51GB |
23
+ | [TinyLlama-1.1B-Chat-v0.5.Q3_K_M.gguf](https://huggingface.co/RichardErkhov/TinyLlama_-_TinyLlama-1.1B-Chat-v0.5-gguf/blob/main/TinyLlama-1.1B-Chat-v0.5.Q3_K_M.gguf) | Q3_K_M | 0.51GB |
24
+ | [TinyLlama-1.1B-Chat-v0.5.Q3_K_L.gguf](https://huggingface.co/RichardErkhov/TinyLlama_-_TinyLlama-1.1B-Chat-v0.5-gguf/blob/main/TinyLlama-1.1B-Chat-v0.5.Q3_K_L.gguf) | Q3_K_L | 0.55GB |
25
+ | [TinyLlama-1.1B-Chat-v0.5.IQ4_XS.gguf](https://huggingface.co/RichardErkhov/TinyLlama_-_TinyLlama-1.1B-Chat-v0.5-gguf/blob/main/TinyLlama-1.1B-Chat-v0.5.IQ4_XS.gguf) | IQ4_XS | 0.57GB |
26
+ | [TinyLlama-1.1B-Chat-v0.5.Q4_0.gguf](https://huggingface.co/RichardErkhov/TinyLlama_-_TinyLlama-1.1B-Chat-v0.5-gguf/blob/main/TinyLlama-1.1B-Chat-v0.5.Q4_0.gguf) | Q4_0 | 0.59GB |
27
+ | [TinyLlama-1.1B-Chat-v0.5.IQ4_NL.gguf](https://huggingface.co/RichardErkhov/TinyLlama_-_TinyLlama-1.1B-Chat-v0.5-gguf/blob/main/TinyLlama-1.1B-Chat-v0.5.IQ4_NL.gguf) | IQ4_NL | 0.6GB |
28
+ | [TinyLlama-1.1B-Chat-v0.5.Q4_K_S.gguf](https://huggingface.co/RichardErkhov/TinyLlama_-_TinyLlama-1.1B-Chat-v0.5-gguf/blob/main/TinyLlama-1.1B-Chat-v0.5.Q4_K_S.gguf) | Q4_K_S | 0.6GB |
29
+ | [TinyLlama-1.1B-Chat-v0.5.Q4_K.gguf](https://huggingface.co/RichardErkhov/TinyLlama_-_TinyLlama-1.1B-Chat-v0.5-gguf/blob/main/TinyLlama-1.1B-Chat-v0.5.Q4_K.gguf) | Q4_K | 0.62GB |
30
+ | [TinyLlama-1.1B-Chat-v0.5.Q4_K_M.gguf](https://huggingface.co/RichardErkhov/TinyLlama_-_TinyLlama-1.1B-Chat-v0.5-gguf/blob/main/TinyLlama-1.1B-Chat-v0.5.Q4_K_M.gguf) | Q4_K_M | 0.62GB |
31
+ | [TinyLlama-1.1B-Chat-v0.5.Q4_1.gguf](https://huggingface.co/RichardErkhov/TinyLlama_-_TinyLlama-1.1B-Chat-v0.5-gguf/blob/main/TinyLlama-1.1B-Chat-v0.5.Q4_1.gguf) | Q4_1 | 0.65GB |
32
+ | [TinyLlama-1.1B-Chat-v0.5.Q5_0.gguf](https://huggingface.co/RichardErkhov/TinyLlama_-_TinyLlama-1.1B-Chat-v0.5-gguf/blob/main/TinyLlama-1.1B-Chat-v0.5.Q5_0.gguf) | Q5_0 | 0.71GB |
33
+ | [TinyLlama-1.1B-Chat-v0.5.Q5_K_S.gguf](https://huggingface.co/RichardErkhov/TinyLlama_-_TinyLlama-1.1B-Chat-v0.5-gguf/blob/main/TinyLlama-1.1B-Chat-v0.5.Q5_K_S.gguf) | Q5_K_S | 0.71GB |
34
+ | [TinyLlama-1.1B-Chat-v0.5.Q5_K.gguf](https://huggingface.co/RichardErkhov/TinyLlama_-_TinyLlama-1.1B-Chat-v0.5-gguf/blob/main/TinyLlama-1.1B-Chat-v0.5.Q5_K.gguf) | Q5_K | 0.73GB |
35
+ | [TinyLlama-1.1B-Chat-v0.5.Q5_K_M.gguf](https://huggingface.co/RichardErkhov/TinyLlama_-_TinyLlama-1.1B-Chat-v0.5-gguf/blob/main/TinyLlama-1.1B-Chat-v0.5.Q5_K_M.gguf) | Q5_K_M | 0.73GB |
36
+ | [TinyLlama-1.1B-Chat-v0.5.Q5_1.gguf](https://huggingface.co/RichardErkhov/TinyLlama_-_TinyLlama-1.1B-Chat-v0.5-gguf/blob/main/TinyLlama-1.1B-Chat-v0.5.Q5_1.gguf) | Q5_1 | 0.77GB |
37
+ | [TinyLlama-1.1B-Chat-v0.5.Q6_K.gguf](https://huggingface.co/RichardErkhov/TinyLlama_-_TinyLlama-1.1B-Chat-v0.5-gguf/blob/main/TinyLlama-1.1B-Chat-v0.5.Q6_K.gguf) | Q6_K | 0.84GB |
38
+ | [TinyLlama-1.1B-Chat-v0.5.Q8_0.gguf](https://huggingface.co/RichardErkhov/TinyLlama_-_TinyLlama-1.1B-Chat-v0.5-gguf/blob/main/TinyLlama-1.1B-Chat-v0.5.Q8_0.gguf) | Q8_0 | 1.09GB |
39
+
40
+
41
+
42
+
43
+ Original model description:
44
+ ---
45
+ license: apache-2.0
46
+ datasets:
47
+ - cerebras/SlimPajama-627B
48
+ - bigcode/starcoderdata
49
+ - OpenAssistant/oasst_top1_2023-08-25
50
+ language:
51
+ - en
52
+ ---
53
+ <div align="center">
54
+
55
+ # TinyLlama-1.1B
56
+ </div>
57
+
58
+ https://github.com/jzhang38/TinyLlama
59
+
60
+ 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 🚀🚀. The training has started on 2023-09-01.
61
+
62
+
63
+ 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.
64
+
65
+ #### This Model
66
+ This is the chat model finetuned on top of [TinyLlama/TinyLlama-1.1B-intermediate-step-955k-2T](https://huggingface.co/TinyLlama/TinyLlama-1.1B-intermediate-step-955k-token-2T).
67
+ The dataset used is [OpenAssistant/oasst_top1_2023-08-25](https://huggingface.co/datasets/OpenAssistant/oasst_top1_2023-08-25) following the [chatml](https://github.com/openai/openai-python/blob/main/chatml.md) format.
68
+ #### How to use
69
+ You will need the transformers>=4.31
70
+ Do check the [TinyLlama](https://github.com/jzhang38/TinyLlama) github page for more information.
71
+ ```
72
+ from transformers import AutoTokenizer
73
+ import transformers
74
+ import torch
75
+ model = "PY007/TinyLlama-1.1B-Chat-v0.5"
76
+ tokenizer = AutoTokenizer.from_pretrained(model)
77
+ pipeline = transformers.pipeline(
78
+ "text-generation",
79
+ model=model,
80
+ torch_dtype=torch.float16,
81
+ device_map="auto",
82
+ )
83
+
84
+ CHAT_EOS_TOKEN_ID = 32002
85
+
86
+ prompt = "How to get in a good university?"
87
+ formatted_prompt = (
88
+ f"<|im_start|>user\n{prompt}<|im_end|>\n<|im_start|>assistant\n"
89
+ )
90
+
91
+
92
+ sequences = pipeline(
93
+ formatted_prompt,
94
+ do_sample=True,
95
+ top_k=50,
96
+ top_p = 0.9,
97
+ num_return_sequences=1,
98
+ repetition_penalty=1.1,
99
+ max_new_tokens=1024,
100
+ eos_token_id=CHAT_EOS_TOKEN_ID,
101
+ )
102
+
103
+ for seq in sequences:
104
+ print(f"Result: {seq['generated_text']}")
105
+ ```
106
+