nassimjp
/

nassimjp artindnr commited on
Commit
fc6aebb
·
0 Parent(s):

Duplicate from artindnr/tea-4bit

Browse files

Co-authored-by: Artin Daneshvar <artindnr@users.noreply.huggingface.co>

.gitattributes ADDED
@@ -0,0 +1,35 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ *.7z filter=lfs diff=lfs merge=lfs -text
2
+ *.arrow filter=lfs diff=lfs merge=lfs -text
3
+ *.bin filter=lfs diff=lfs merge=lfs -text
4
+ *.bz2 filter=lfs diff=lfs merge=lfs -text
5
+ *.ckpt filter=lfs diff=lfs merge=lfs -text
6
+ *.ftz filter=lfs diff=lfs merge=lfs -text
7
+ *.gz filter=lfs diff=lfs merge=lfs -text
8
+ *.h5 filter=lfs diff=lfs merge=lfs -text
9
+ *.joblib filter=lfs diff=lfs merge=lfs -text
10
+ *.lfs.* filter=lfs diff=lfs merge=lfs -text
11
+ *.mlmodel filter=lfs diff=lfs merge=lfs -text
12
+ *.model filter=lfs diff=lfs merge=lfs -text
13
+ *.msgpack filter=lfs diff=lfs merge=lfs -text
14
+ *.npy filter=lfs diff=lfs merge=lfs -text
15
+ *.npz filter=lfs diff=lfs merge=lfs -text
16
+ *.onnx filter=lfs diff=lfs merge=lfs -text
17
+ *.ot filter=lfs diff=lfs merge=lfs -text
18
+ *.parquet filter=lfs diff=lfs merge=lfs -text
19
+ *.pb filter=lfs diff=lfs merge=lfs -text
20
+ *.pickle filter=lfs diff=lfs merge=lfs -text
21
+ *.pkl filter=lfs diff=lfs merge=lfs -text
22
+ *.pt filter=lfs diff=lfs merge=lfs -text
23
+ *.pth filter=lfs diff=lfs merge=lfs -text
24
+ *.rar filter=lfs diff=lfs merge=lfs -text
25
+ *.safetensors filter=lfs diff=lfs merge=lfs -text
26
+ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
27
+ *.tar.* filter=lfs diff=lfs merge=lfs -text
28
+ *.tar filter=lfs diff=lfs merge=lfs -text
29
+ *.tflite filter=lfs diff=lfs merge=lfs -text
30
+ *.tgz filter=lfs diff=lfs merge=lfs -text
31
+ *.wasm filter=lfs diff=lfs merge=lfs -text
32
+ *.xz filter=lfs diff=lfs merge=lfs -text
33
+ *.zip filter=lfs diff=lfs merge=lfs -text
34
+ *.zst filter=lfs diff=lfs merge=lfs -text
35
+ *tfevents* filter=lfs diff=lfs merge=lfs -text
README.md ADDED
@@ -0,0 +1,152 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: mit
3
+ base_model: artindnr/tea
4
+ tags:
5
+ - fine-tuned
6
+ - full-fine-tune
7
+ - 4-bit
8
+ - bitsandbytes
9
+ - unsloth
10
+ - safetensors
11
+ - text-generation
12
+ - chat
13
+ - question-answering
14
+ - assistant
15
+ language:
16
+ - fa
17
+ - en
18
+ - multilingual
19
+ pipeline_tag: text-generation
20
+ ---
21
+
22
+ # 🍵 Tea — 4-bit
23
+ ![https://64.media.tumblr.com/1ab2cfe03429ef47b5063c90df2c0a5a/3c1a235d6f992b74-7d/s500x750/aafa42b7949f3104f349e5508df5fa8b738d879a.gif](https://64.media.tumblr.com/1ab2cfe03429ef47b5063c90df2c0a5a/3c1a235d6f992b74-7d/s500x750/aafa42b7949f3104f349e5508df5fa8b738d879a.gif)
24
+
25
+ This repo contains a **4-bit precision version of [`artindnr/tea`](https://huggingface.co/artindnr/tea)**, a full fine-tune of [`microsoft/phi-4`](https://huggingface.co/microsoft/phi-4) for question answering and long, multi-turn assistant conversations, with fine-tuning focused on Farsi (Persian) conversational ability.
26
+
27
+ Unlike GGUF quants, this is a **safetensors** checkpoint quantized to 4-bit precision with [Unsloth](https://github.com/unslothai/unsloth) (bitsandbytes nf4 backend), meant to be loaded directly with 🤗 Transformers or Unsloth — not with llama.cpp.
28
+
29
+ > Looking for GGUF quants for llama.cpp / Ollama / LM Studio instead? See [artindnr/tea-gguf](https://huggingface.co/artindnr/tea-gguf) (F16, Q8_0, Q5_K_M, Q4_K_M).
30
+
31
+ ## Model Details
32
+
33
+ - **Base model:** [artindnr/tea](https://huggingface.co/artindnr/tea) (full fine-tune of `microsoft/phi-4`, 14B parameters)
34
+ - **Quantized by:** [artindnr](https://huggingface.co/artindnr), using [Unsloth](https://github.com/unslothai/unsloth)
35
+ - **Format:** safetensors (4-bit, bitsandbytes nf4)
36
+ - **License:** MIT
37
+ - **Languages:** Farsi (primary conversational focus), English, and general multilingual support
38
+
39
+ ## How to Use
40
+
41
+ ### Generation with 🤗 Transformers
42
+
43
+ The checkpoint is already stored in 4-bit, so it loads directly — no `BitsAndBytesConfig` needed on your end, `device_map="auto"` handles placement.
44
+
45
+ ```python
46
+ import torch
47
+ from transformers import AutoModelForCausalLM, AutoTokenizer
48
+
49
+ MODEL_ID = "artindnr/tea-4bit"
50
+
51
+ tokenizer = AutoTokenizer.from_pretrained(MODEL_ID)
52
+ model = AutoModelForCausalLM.from_pretrained(
53
+ MODEL_ID,
54
+ device_map="auto",
55
+ )
56
+
57
+ USER_PROMPT = "تو کی هستی و اسمت چیه؟"
58
+
59
+ messages = [
60
+ {"role": "user", "content": USER_PROMPT},
61
+ ]
62
+
63
+ inputs = tokenizer.apply_chat_template(
64
+ messages,
65
+ add_generation_prompt=True,
66
+ tokenize=True,
67
+ return_dict=True,
68
+ return_tensors="pt",
69
+ ).to(model.device)
70
+
71
+ outputs = model.generate(
72
+ **inputs,
73
+ max_new_tokens=1024,
74
+ temperature=0.7,
75
+ do_sample=True,
76
+ )
77
+
78
+ print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:], skip_special_tokens=True))
79
+ ```
80
+
81
+ ### Generation with Unsloth (faster inference/fine-tuning)
82
+
83
+ ```bash
84
+ pip install unsloth
85
+ ```
86
+
87
+ ```python
88
+ from unsloth import FastLanguageModel
89
+
90
+ model, tokenizer = FastLanguageModel.from_pretrained(
91
+ model_name="artindnr/tea-4bit",
92
+ max_seq_length=8192,
93
+ load_in_4bit=True,
94
+ dtype=None, # auto-detect
95
+ )
96
+ FastLanguageModel.for_inference(model)
97
+
98
+ messages = [
99
+ {"role": "user", "content": "تو کی هستی و اسمت چیه؟"},
100
+ ]
101
+
102
+ inputs = tokenizer.apply_chat_template(
103
+ messages,
104
+ add_generation_prompt=True,
105
+ tokenize=True,
106
+ return_dict=True,
107
+ return_tensors="pt",
108
+ ).to(model.device)
109
+
110
+ outputs = model.generate(
111
+ **inputs,
112
+ max_new_tokens=1024,
113
+ temperature=0.7,
114
+ do_sample=True,
115
+ )
116
+
117
+ print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:], skip_special_tokens=True))
118
+ ```
119
+
120
+ Unsloth also lets you use this checkpoint as a starting point for further QLoRA fine-tuning if you want to adapt tea further while keeping it in 4-bit.
121
+
122
+ ## Intended Use
123
+
124
+ Same as [`artindnr/tea`](https://huggingface.co/artindnr/tea): Farsi-first conversational assistance, question answering, and long multi-turn assistant deployments — this repo specifically targets **lower-VRAM GPU inference and fine-tuning** via Transformers/Unsloth, as an alternative to the GGUF quants for CPU/llama.cpp-based deployment.
125
+
126
+ ## Limitations
127
+
128
+ - 4-bit quantization trades off some accuracy for memory footprint; expect some quality degradation versus the full-precision [`artindnr/tea`](https://huggingface.co/artindnr/tea), particularly on nuanced or long-context Farsi generation.
129
+ - This checkpoint requires a CUDA GPU and `bitsandbytes` — it is not intended for CPU inference or llama.cpp; use [artindnr/tea-gguf](https://huggingface.co/artindnr/tea-gguf) for that.
130
+ - Inherits all limitations of the base [`artindnr/tea`](https://huggingface.co/artindnr/tea) model and the underlying `microsoft/phi-4` checkpoint, including possible hallucinated facts.
131
+ - No formal safety fine-tuning beyond what is inherited from the base model has been applied; use appropriate safeguards in production settings.
132
+
133
+ ## License
134
+
135
+ This model is released under the [MIT License](https://opensource.org/licenses/MIT), consistent with `artindnr/tea` and the base `microsoft/phi-4` model.
136
+
137
+ ## Citation
138
+
139
+ If you use tea in your work, please cite:
140
+
141
+ ```bibtex
142
+ @misc{tea,
143
+ title = {tea: A Farsi-Focused, Full Fine-tune of Phi-4 for QA and Long-form Assistance},
144
+ author = {artindnr},
145
+ year = {2026},
146
+ url = {https://huggingface.co/artindnr/tea}
147
+ }
148
+ ```
149
+
150
+ ## Acknowledgements
151
+
152
+ Built on top of [`artindnr/tea`](https://huggingface.co/artindnr/tea), itself a full fine-tune of [`microsoft/phi-4`](https://huggingface.co/microsoft/phi-4). 4-bit quantization via [Unsloth](https://github.com/unslothai/unsloth).
chat_template.jinja ADDED
@@ -0,0 +1 @@
 
 
1
+ {% for message in messages %}{% if (message['role'] == 'system') %}{{'<|im_start|>system<|im_sep|>' + message['content'] + '<|im_end|>'}}{% elif (message['role'] == 'user') %}{{'<|im_start|>user<|im_sep|>' + message['content'] + '<|im_end|>'}}{% elif (message['role'] == 'assistant') %}{{'<|im_start|>assistant<|im_sep|>' + message['content'] + '<|im_end|>'}}{% endif %}{% endfor %}{% if add_generation_prompt %}{{ '<|im_start|>assistant<|im_sep|>' }}{% endif %}
config.json ADDED
@@ -0,0 +1,58 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "LlamaForCausalLM"
4
+ ],
5
+ "attention_bias": false,
6
+ "attention_dropout": 0.0,
7
+ "bos_token_id": 100257,
8
+ "torch_dtype": "bfloat16",
9
+ "eos_token_id": 100265,
10
+ "head_dim": 128,
11
+ "hidden_act": "silu",
12
+ "hidden_size": 5120,
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 17920,
15
+ "max_position_embeddings": 16384,
16
+ "mlp_bias": false,
17
+ "model_type": "llama",
18
+ "num_attention_heads": 40,
19
+ "num_hidden_layers": 40,
20
+ "num_key_value_heads": 10,
21
+ "original_max_position_embeddings": 16384,
22
+ "pad_token_id": 100351,
23
+ "pretraining_tp": 1,
24
+ "quantization_config": {
25
+ "bnb_4bit_compute_dtype": "bfloat16",
26
+ "bnb_4bit_quant_type": "nf4",
27
+ "bnb_4bit_use_double_quant": true,
28
+ "llm_int8_enable_fp32_cpu_offload": false,
29
+ "llm_int8_has_fp16_weight": false,
30
+ "llm_int8_skip_modules": [
31
+ "model.layers.1.mlp.gate_proj",
32
+ "model.layers.1.mlp.up_proj",
33
+ "model.layers.1.mlp.down_proj",
34
+ "model.layers.2.mlp.gate_proj",
35
+ "model.layers.2.mlp.up_proj",
36
+ "model.layers.2.mlp.down_proj",
37
+ "model.layers.3.mlp.gate_proj",
38
+ "model.layers.3.mlp.up_proj",
39
+ "model.layers.3.mlp.down_proj",
40
+ "model.layers.37.self_attn.o_proj",
41
+ "model.layers.39.self_attn.o_proj",
42
+ "lm_head"
43
+ ],
44
+ "llm_int8_threshold": 6.0,
45
+ "load_in_4bit": true,
46
+ "load_in_8bit": false,
47
+ "quant_method": "bitsandbytes"
48
+ },
49
+ "rms_norm_eps": 1e-05,
50
+ "rope_scaling": null,
51
+ "rope_theta": 250000,
52
+ "tie_word_embeddings": false,
53
+ "transformers_version": "4.56.2",
54
+ "unsloth_fixed": true,
55
+ "unsloth_version": "2026.8.8",
56
+ "use_cache": true,
57
+ "vocab_size": 100352
58
+ }
generation_config.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 100257,
4
+ "eos_token_id": [
5
+ 100265
6
+ ],
7
+ "max_length": 16384,
8
+ "pad_token_id": 100351,
9
+ "transformers_version": "4.56.2"
10
+ }
merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
model-00001-of-00003.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:78e2e25a5317d0017dfd0130ad81cbd7d44807b850ac593ad329b2ca2b319162
3
+ size 4971805405
model-00002-of-00003.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:85ca66046711888e091e7272a4454c1834e7d53f38d7c7bc81418239078ab3af
3
+ size 4392572573
model-00003-of-00003.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5839c81af91221e4d409b7fac2c3e7d348774887a39841e184ac73bc68e7c1f3
3
+ size 1027604608
model.safetensors.index.json ADDED
The diff for this file is too large to render. See raw diff
 
special_tokens_map.json ADDED
@@ -0,0 +1,30 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": {
3
+ "content": "<|endoftext|>",
4
+ "lstrip": true,
5
+ "normalized": false,
6
+ "rstrip": true,
7
+ "single_word": false
8
+ },
9
+ "eos_token": {
10
+ "content": "<|im_end|>",
11
+ "lstrip": true,
12
+ "normalized": false,
13
+ "rstrip": true,
14
+ "single_word": false
15
+ },
16
+ "pad_token": {
17
+ "content": "<|dummy_87|>",
18
+ "lstrip": true,
19
+ "normalized": false,
20
+ "rstrip": true,
21
+ "single_word": false
22
+ },
23
+ "unk_token": {
24
+ "content": "�",
25
+ "lstrip": false,
26
+ "normalized": false,
27
+ "rstrip": false,
28
+ "single_word": false
29
+ }
30
+ }
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer_config.json ADDED
@@ -0,0 +1,791 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "added_tokens_decoder": {
4
+ "5809": {
5
+ "content": "�",
6
+ "lstrip": false,
7
+ "normalized": false,
8
+ "rstrip": false,
9
+ "single_word": false,
10
+ "special": true
11
+ },
12
+ "100256": {
13
+ "content": "<|dummy_0|>",
14
+ "lstrip": true,
15
+ "normalized": false,
16
+ "rstrip": true,
17
+ "single_word": false,
18
+ "special": true
19
+ },
20
+ "100257": {
21
+ "content": "<|endoftext|>",
22
+ "lstrip": true,
23
+ "normalized": false,
24
+ "rstrip": true,
25
+ "single_word": false,
26
+ "special": true
27
+ },
28
+ "100258": {
29
+ "content": "<|fim_prefix|>",
30
+ "lstrip": true,
31
+ "normalized": false,
32
+ "rstrip": true,
33
+ "single_word": false,
34
+ "special": true
35
+ },
36
+ "100259": {
37
+ "content": "<|fim_middle|>",
38
+ "lstrip": true,
39
+ "normalized": false,
40
+ "rstrip": true,
41
+ "single_word": false,
42
+ "special": true
43
+ },
44
+ "100260": {
45
+ "content": "<|fim_suffix|>",
46
+ "lstrip": true,
47
+ "normalized": false,
48
+ "rstrip": true,
49
+ "single_word": false,
50
+ "special": true
51
+ },
52
+ "100261": {
53
+ "content": "<|dummy_1|>",
54
+ "lstrip": true,
55
+ "normalized": false,
56
+ "rstrip": true,
57
+ "single_word": false,
58
+ "special": true
59
+ },
60
+ "100262": {
61
+ "content": "<|dummy_2|>",
62
+ "lstrip": true,
63
+ "normalized": false,
64
+ "rstrip": true,
65
+ "single_word": false,
66
+ "special": true
67
+ },
68
+ "100263": {
69
+ "content": "<|dummy_3|>",
70
+ "lstrip": true,
71
+ "normalized": false,
72
+ "rstrip": true,
73
+ "single_word": false,
74
+ "special": true
75
+ },
76
+ "100264": {
77
+ "content": "<|im_start|>",
78
+ "lstrip": true,
79
+ "normalized": false,
80
+ "rstrip": true,
81
+ "single_word": false,
82
+ "special": true
83
+ },
84
+ "100265": {
85
+ "content": "<|im_end|>",
86
+ "lstrip": true,
87
+ "normalized": false,
88
+ "rstrip": true,
89
+ "single_word": false,
90
+ "special": true
91
+ },
92
+ "100266": {
93
+ "content": "<|im_sep|>",
94
+ "lstrip": true,
95
+ "normalized": false,
96
+ "rstrip": true,
97
+ "single_word": false,
98
+ "special": true
99
+ },
100
+ "100267": {
101
+ "content": "<|dummy_4|>",
102
+ "lstrip": true,
103
+ "normalized": false,
104
+ "rstrip": true,
105
+ "single_word": false,
106
+ "special": true
107
+ },
108
+ "100268": {
109
+ "content": "<|dummy_5|>",
110
+ "lstrip": true,
111
+ "normalized": false,
112
+ "rstrip": true,
113
+ "single_word": false,
114
+ "special": true
115
+ },
116
+ "100269": {
117
+ "content": "<|dummy_6|>",
118
+ "lstrip": true,
119
+ "normalized": false,
120
+ "rstrip": true,
121
+ "single_word": false,
122
+ "special": true
123
+ },
124
+ "100270": {
125
+ "content": "<|dummy_7|>",
126
+ "lstrip": true,
127
+ "normalized": false,
128
+ "rstrip": true,
129
+ "single_word": false,
130
+ "special": true
131
+ },
132
+ "100271": {
133
+ "content": "<|dummy_8|>",
134
+ "lstrip": true,
135
+ "normalized": false,
136
+ "rstrip": true,
137
+ "single_word": false,
138
+ "special": true
139
+ },
140
+ "100272": {
141
+ "content": "<|dummy_9|>",
142
+ "lstrip": true,
143
+ "normalized": false,
144
+ "rstrip": true,
145
+ "single_word": false,
146
+ "special": true
147
+ },
148
+ "100273": {
149
+ "content": "<|dummy_10|>",
150
+ "lstrip": true,
151
+ "normalized": false,
152
+ "rstrip": true,
153
+ "single_word": false,
154
+ "special": true
155
+ },
156
+ "100274": {
157
+ "content": "<|dummy_11|>",
158
+ "lstrip": true,
159
+ "normalized": false,
160
+ "rstrip": true,
161
+ "single_word": false,
162
+ "special": true
163
+ },
164
+ "100275": {
165
+ "content": "<|dummy_12|>",
166
+ "lstrip": true,
167
+ "normalized": false,
168
+ "rstrip": true,
169
+ "single_word": false,
170
+ "special": true
171
+ },
172
+ "100276": {
173
+ "content": "<|endofprompt|>",
174
+ "lstrip": true,
175
+ "normalized": false,
176
+ "rstrip": true,
177
+ "single_word": false,
178
+ "special": true
179
+ },
180
+ "100277": {
181
+ "content": "<|dummy_13|>",
182
+ "lstrip": true,
183
+ "normalized": false,
184
+ "rstrip": true,
185
+ "single_word": false,
186
+ "special": true
187
+ },
188
+ "100278": {
189
+ "content": "<|dummy_14|>",
190
+ "lstrip": true,
191
+ "normalized": false,
192
+ "rstrip": true,
193
+ "single_word": false,
194
+ "special": true
195
+ },
196
+ "100279": {
197
+ "content": "<|dummy_15|>",
198
+ "lstrip": true,
199
+ "normalized": false,
200
+ "rstrip": true,
201
+ "single_word": false,
202
+ "special": true
203
+ },
204
+ "100280": {
205
+ "content": "<|dummy_16|>",
206
+ "lstrip": true,
207
+ "normalized": false,
208
+ "rstrip": true,
209
+ "single_word": false,
210
+ "special": true
211
+ },
212
+ "100281": {
213
+ "content": "<|dummy_17|>",
214
+ "lstrip": true,
215
+ "normalized": false,
216
+ "rstrip": true,
217
+ "single_word": false,
218
+ "special": true
219
+ },
220
+ "100282": {
221
+ "content": "<|dummy_18|>",
222
+ "lstrip": true,
223
+ "normalized": false,
224
+ "rstrip": true,
225
+ "single_word": false,
226
+ "special": true
227
+ },
228
+ "100283": {
229
+ "content": "<|dummy_19|>",
230
+ "lstrip": true,
231
+ "normalized": false,
232
+ "rstrip": true,
233
+ "single_word": false,
234
+ "special": true
235
+ },
236
+ "100284": {
237
+ "content": "<|dummy_20|>",
238
+ "lstrip": true,
239
+ "normalized": false,
240
+ "rstrip": true,
241
+ "single_word": false,
242
+ "special": true
243
+ },
244
+ "100285": {
245
+ "content": "<|dummy_21|>",
246
+ "lstrip": true,
247
+ "normalized": false,
248
+ "rstrip": true,
249
+ "single_word": false,
250
+ "special": true
251
+ },
252
+ "100286": {
253
+ "content": "<|dummy_22|>",
254
+ "lstrip": true,
255
+ "normalized": false,
256
+ "rstrip": true,
257
+ "single_word": false,
258
+ "special": true
259
+ },
260
+ "100287": {
261
+ "content": "<|dummy_23|>",
262
+ "lstrip": true,
263
+ "normalized": false,
264
+ "rstrip": true,
265
+ "single_word": false,
266
+ "special": true
267
+ },
268
+ "100288": {
269
+ "content": "<|dummy_24|>",
270
+ "lstrip": true,
271
+ "normalized": false,
272
+ "rstrip": true,
273
+ "single_word": false,
274
+ "special": true
275
+ },
276
+ "100289": {
277
+ "content": "<|dummy_25|>",
278
+ "lstrip": true,
279
+ "normalized": false,
280
+ "rstrip": true,
281
+ "single_word": false,
282
+ "special": true
283
+ },
284
+ "100290": {
285
+ "content": "<|dummy_26|>",
286
+ "lstrip": true,
287
+ "normalized": false,
288
+ "rstrip": true,
289
+ "single_word": false,
290
+ "special": true
291
+ },
292
+ "100291": {
293
+ "content": "<|dummy_27|>",
294
+ "lstrip": true,
295
+ "normalized": false,
296
+ "rstrip": true,
297
+ "single_word": false,
298
+ "special": true
299
+ },
300
+ "100292": {
301
+ "content": "<|dummy_28|>",
302
+ "lstrip": true,
303
+ "normalized": false,
304
+ "rstrip": true,
305
+ "single_word": false,
306
+ "special": true
307
+ },
308
+ "100293": {
309
+ "content": "<|dummy_29|>",
310
+ "lstrip": true,
311
+ "normalized": false,
312
+ "rstrip": true,
313
+ "single_word": false,
314
+ "special": true
315
+ },
316
+ "100294": {
317
+ "content": "<|dummy_30|>",
318
+ "lstrip": true,
319
+ "normalized": false,
320
+ "rstrip": true,
321
+ "single_word": false,
322
+ "special": true
323
+ },
324
+ "100295": {
325
+ "content": "<|dummy_31|>",
326
+ "lstrip": true,
327
+ "normalized": false,
328
+ "rstrip": true,
329
+ "single_word": false,
330
+ "special": true
331
+ },
332
+ "100296": {
333
+ "content": "<|dummy_32|>",
334
+ "lstrip": true,
335
+ "normalized": false,
336
+ "rstrip": true,
337
+ "single_word": false,
338
+ "special": true
339
+ },
340
+ "100297": {
341
+ "content": "<|dummy_33|>",
342
+ "lstrip": true,
343
+ "normalized": false,
344
+ "rstrip": true,
345
+ "single_word": false,
346
+ "special": true
347
+ },
348
+ "100298": {
349
+ "content": "<|dummy_34|>",
350
+ "lstrip": true,
351
+ "normalized": false,
352
+ "rstrip": true,
353
+ "single_word": false,
354
+ "special": true
355
+ },
356
+ "100299": {
357
+ "content": "<|dummy_35|>",
358
+ "lstrip": true,
359
+ "normalized": false,
360
+ "rstrip": true,
361
+ "single_word": false,
362
+ "special": true
363
+ },
364
+ "100300": {
365
+ "content": "<|dummy_36|>",
366
+ "lstrip": true,
367
+ "normalized": false,
368
+ "rstrip": true,
369
+ "single_word": false,
370
+ "special": true
371
+ },
372
+ "100301": {
373
+ "content": "<|dummy_37|>",
374
+ "lstrip": true,
375
+ "normalized": false,
376
+ "rstrip": true,
377
+ "single_word": false,
378
+ "special": true
379
+ },
380
+ "100302": {
381
+ "content": "<|dummy_38|>",
382
+ "lstrip": true,
383
+ "normalized": false,
384
+ "rstrip": true,
385
+ "single_word": false,
386
+ "special": true
387
+ },
388
+ "100303": {
389
+ "content": "<|dummy_39|>",
390
+ "lstrip": true,
391
+ "normalized": false,
392
+ "rstrip": true,
393
+ "single_word": false,
394
+ "special": true
395
+ },
396
+ "100304": {
397
+ "content": "<|dummy_40|>",
398
+ "lstrip": true,
399
+ "normalized": false,
400
+ "rstrip": true,
401
+ "single_word": false,
402
+ "special": true
403
+ },
404
+ "100305": {
405
+ "content": "<|dummy_41|>",
406
+ "lstrip": true,
407
+ "normalized": false,
408
+ "rstrip": true,
409
+ "single_word": false,
410
+ "special": true
411
+ },
412
+ "100306": {
413
+ "content": "<|dummy_42|>",
414
+ "lstrip": true,
415
+ "normalized": false,
416
+ "rstrip": true,
417
+ "single_word": false,
418
+ "special": true
419
+ },
420
+ "100307": {
421
+ "content": "<|dummy_43|>",
422
+ "lstrip": true,
423
+ "normalized": false,
424
+ "rstrip": true,
425
+ "single_word": false,
426
+ "special": true
427
+ },
428
+ "100308": {
429
+ "content": "<|dummy_44|>",
430
+ "lstrip": true,
431
+ "normalized": false,
432
+ "rstrip": true,
433
+ "single_word": false,
434
+ "special": true
435
+ },
436
+ "100309": {
437
+ "content": "<|dummy_45|>",
438
+ "lstrip": true,
439
+ "normalized": false,
440
+ "rstrip": true,
441
+ "single_word": false,
442
+ "special": true
443
+ },
444
+ "100310": {
445
+ "content": "<|dummy_46|>",
446
+ "lstrip": true,
447
+ "normalized": false,
448
+ "rstrip": true,
449
+ "single_word": false,
450
+ "special": true
451
+ },
452
+ "100311": {
453
+ "content": "<|dummy_47|>",
454
+ "lstrip": true,
455
+ "normalized": false,
456
+ "rstrip": true,
457
+ "single_word": false,
458
+ "special": true
459
+ },
460
+ "100312": {
461
+ "content": "<|dummy_48|>",
462
+ "lstrip": true,
463
+ "normalized": false,
464
+ "rstrip": true,
465
+ "single_word": false,
466
+ "special": true
467
+ },
468
+ "100313": {
469
+ "content": "<|dummy_49|>",
470
+ "lstrip": true,
471
+ "normalized": false,
472
+ "rstrip": true,
473
+ "single_word": false,
474
+ "special": true
475
+ },
476
+ "100314": {
477
+ "content": "<|dummy_50|>",
478
+ "lstrip": true,
479
+ "normalized": false,
480
+ "rstrip": true,
481
+ "single_word": false,
482
+ "special": true
483
+ },
484
+ "100315": {
485
+ "content": "<|dummy_51|>",
486
+ "lstrip": true,
487
+ "normalized": false,
488
+ "rstrip": true,
489
+ "single_word": false,
490
+ "special": true
491
+ },
492
+ "100316": {
493
+ "content": "<|dummy_52|>",
494
+ "lstrip": true,
495
+ "normalized": false,
496
+ "rstrip": true,
497
+ "single_word": false,
498
+ "special": true
499
+ },
500
+ "100317": {
501
+ "content": "<|dummy_53|>",
502
+ "lstrip": true,
503
+ "normalized": false,
504
+ "rstrip": true,
505
+ "single_word": false,
506
+ "special": true
507
+ },
508
+ "100318": {
509
+ "content": "<|dummy_54|>",
510
+ "lstrip": true,
511
+ "normalized": false,
512
+ "rstrip": true,
513
+ "single_word": false,
514
+ "special": true
515
+ },
516
+ "100319": {
517
+ "content": "<|dummy_55|>",
518
+ "lstrip": true,
519
+ "normalized": false,
520
+ "rstrip": true,
521
+ "single_word": false,
522
+ "special": true
523
+ },
524
+ "100320": {
525
+ "content": "<|dummy_56|>",
526
+ "lstrip": true,
527
+ "normalized": false,
528
+ "rstrip": true,
529
+ "single_word": false,
530
+ "special": true
531
+ },
532
+ "100321": {
533
+ "content": "<|dummy_57|>",
534
+ "lstrip": true,
535
+ "normalized": false,
536
+ "rstrip": true,
537
+ "single_word": false,
538
+ "special": true
539
+ },
540
+ "100322": {
541
+ "content": "<|dummy_58|>",
542
+ "lstrip": true,
543
+ "normalized": false,
544
+ "rstrip": true,
545
+ "single_word": false,
546
+ "special": true
547
+ },
548
+ "100323": {
549
+ "content": "<|dummy_59|>",
550
+ "lstrip": true,
551
+ "normalized": false,
552
+ "rstrip": true,
553
+ "single_word": false,
554
+ "special": true
555
+ },
556
+ "100324": {
557
+ "content": "<|dummy_60|>",
558
+ "lstrip": true,
559
+ "normalized": false,
560
+ "rstrip": true,
561
+ "single_word": false,
562
+ "special": true
563
+ },
564
+ "100325": {
565
+ "content": "<|dummy_61|>",
566
+ "lstrip": true,
567
+ "normalized": false,
568
+ "rstrip": true,
569
+ "single_word": false,
570
+ "special": true
571
+ },
572
+ "100326": {
573
+ "content": "<|dummy_62|>",
574
+ "lstrip": true,
575
+ "normalized": false,
576
+ "rstrip": true,
577
+ "single_word": false,
578
+ "special": true
579
+ },
580
+ "100327": {
581
+ "content": "<|dummy_63|>",
582
+ "lstrip": true,
583
+ "normalized": false,
584
+ "rstrip": true,
585
+ "single_word": false,
586
+ "special": true
587
+ },
588
+ "100328": {
589
+ "content": "<|dummy_64|>",
590
+ "lstrip": true,
591
+ "normalized": false,
592
+ "rstrip": true,
593
+ "single_word": false,
594
+ "special": true
595
+ },
596
+ "100329": {
597
+ "content": "<|dummy_65|>",
598
+ "lstrip": true,
599
+ "normalized": false,
600
+ "rstrip": true,
601
+ "single_word": false,
602
+ "special": true
603
+ },
604
+ "100330": {
605
+ "content": "<|dummy_66|>",
606
+ "lstrip": true,
607
+ "normalized": false,
608
+ "rstrip": true,
609
+ "single_word": false,
610
+ "special": true
611
+ },
612
+ "100331": {
613
+ "content": "<|dummy_67|>",
614
+ "lstrip": true,
615
+ "normalized": false,
616
+ "rstrip": true,
617
+ "single_word": false,
618
+ "special": true
619
+ },
620
+ "100332": {
621
+ "content": "<|dummy_68|>",
622
+ "lstrip": true,
623
+ "normalized": false,
624
+ "rstrip": true,
625
+ "single_word": false,
626
+ "special": true
627
+ },
628
+ "100333": {
629
+ "content": "<|dummy_69|>",
630
+ "lstrip": true,
631
+ "normalized": false,
632
+ "rstrip": true,
633
+ "single_word": false,
634
+ "special": true
635
+ },
636
+ "100334": {
637
+ "content": "<|dummy_70|>",
638
+ "lstrip": true,
639
+ "normalized": false,
640
+ "rstrip": true,
641
+ "single_word": false,
642
+ "special": true
643
+ },
644
+ "100335": {
645
+ "content": "<|dummy_71|>",
646
+ "lstrip": true,
647
+ "normalized": false,
648
+ "rstrip": true,
649
+ "single_word": false,
650
+ "special": true
651
+ },
652
+ "100336": {
653
+ "content": "<|dummy_72|>",
654
+ "lstrip": true,
655
+ "normalized": false,
656
+ "rstrip": true,
657
+ "single_word": false,
658
+ "special": true
659
+ },
660
+ "100337": {
661
+ "content": "<|dummy_73|>",
662
+ "lstrip": true,
663
+ "normalized": false,
664
+ "rstrip": true,
665
+ "single_word": false,
666
+ "special": true
667
+ },
668
+ "100338": {
669
+ "content": "<|dummy_74|>",
670
+ "lstrip": true,
671
+ "normalized": false,
672
+ "rstrip": true,
673
+ "single_word": false,
674
+ "special": true
675
+ },
676
+ "100339": {
677
+ "content": "<|dummy_75|>",
678
+ "lstrip": true,
679
+ "normalized": false,
680
+ "rstrip": true,
681
+ "single_word": false,
682
+ "special": true
683
+ },
684
+ "100340": {
685
+ "content": "<|dummy_76|>",
686
+ "lstrip": true,
687
+ "normalized": false,
688
+ "rstrip": true,
689
+ "single_word": false,
690
+ "special": true
691
+ },
692
+ "100341": {
693
+ "content": "<|dummy_77|>",
694
+ "lstrip": true,
695
+ "normalized": false,
696
+ "rstrip": true,
697
+ "single_word": false,
698
+ "special": true
699
+ },
700
+ "100342": {
701
+ "content": "<|dummy_78|>",
702
+ "lstrip": true,
703
+ "normalized": false,
704
+ "rstrip": true,
705
+ "single_word": false,
706
+ "special": true
707
+ },
708
+ "100343": {
709
+ "content": "<|dummy_79|>",
710
+ "lstrip": true,
711
+ "normalized": false,
712
+ "rstrip": true,
713
+ "single_word": false,
714
+ "special": true
715
+ },
716
+ "100344": {
717
+ "content": "<|dummy_80|>",
718
+ "lstrip": true,
719
+ "normalized": false,
720
+ "rstrip": true,
721
+ "single_word": false,
722
+ "special": true
723
+ },
724
+ "100345": {
725
+ "content": "<|dummy_81|>",
726
+ "lstrip": true,
727
+ "normalized": false,
728
+ "rstrip": true,
729
+ "single_word": false,
730
+ "special": true
731
+ },
732
+ "100346": {
733
+ "content": "<|dummy_82|>",
734
+ "lstrip": true,
735
+ "normalized": false,
736
+ "rstrip": true,
737
+ "single_word": false,
738
+ "special": true
739
+ },
740
+ "100347": {
741
+ "content": "<|dummy_83|>",
742
+ "lstrip": true,
743
+ "normalized": false,
744
+ "rstrip": true,
745
+ "single_word": false,
746
+ "special": true
747
+ },
748
+ "100348": {
749
+ "content": "<|dummy_84|>",
750
+ "lstrip": true,
751
+ "normalized": false,
752
+ "rstrip": true,
753
+ "single_word": false,
754
+ "special": true
755
+ },
756
+ "100349": {
757
+ "content": "<|dummy_85|>",
758
+ "lstrip": true,
759
+ "normalized": false,
760
+ "rstrip": true,
761
+ "single_word": false,
762
+ "special": true
763
+ },
764
+ "100350": {
765
+ "content": "<|dummy_86|>",
766
+ "lstrip": true,
767
+ "normalized": false,
768
+ "rstrip": true,
769
+ "single_word": false,
770
+ "special": true
771
+ },
772
+ "100351": {
773
+ "content": "<|dummy_87|>",
774
+ "lstrip": true,
775
+ "normalized": false,
776
+ "rstrip": true,
777
+ "single_word": false,
778
+ "special": true
779
+ }
780
+ },
781
+ "bos_token": "<|endoftext|>",
782
+ "clean_up_tokenization_spaces": false,
783
+ "eos_token": "<|im_end|>",
784
+ "extra_special_tokens": {},
785
+ "model_max_length": 16384,
786
+ "pad_token": "<|dummy_87|>",
787
+ "padding_side": "left",
788
+ "tokenizer_class": "GPT2Tokenizer",
789
+ "unk_token": "�",
790
+ "chat_template": "{% for message in messages %}{% if (message['role'] == 'system') %}{{'<|im_start|>system<|im_sep|>' + message['content'] + '<|im_end|>'}}{% elif (message['role'] == 'user') %}{{'<|im_start|>user<|im_sep|>' + message['content'] + '<|im_end|>'}}{% elif (message['role'] == 'assistant') %}{{'<|im_start|>assistant<|im_sep|>' + message['content'] + '<|im_end|>'}}{% endif %}{% endfor %}{% if add_generation_prompt %}{{ '<|im_start|>assistant<|im_sep|>' }}{% endif %}"
791
+ }
vocab.json ADDED
The diff for this file is too large to render. See raw diff