puneeshkhanna commited on
Commit
9bde8a9
·
0 Parent(s):

Initial commit

Browse files
.gitattributes ADDED
@@ -0,0 +1,37 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
36
+ tokenizer.json filter=lfs diff=lfs merge=lfs -text
37
+
README.md ADDED
@@ -0,0 +1,179 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ base_model:
3
+ - tiiuae/Falcon-H1R-7B
4
+ language:
5
+ - en
6
+ library_name: transformers
7
+ pipeline_tag: text-generation
8
+ license: other
9
+ license_name: falcon-llm-license
10
+ license_link: https://falconllm.tii.ae/falcon-terms-and-conditions.html
11
+ tags:
12
+ - falcon-h1r
13
+ ---
14
+
15
+ <img src="https://huggingface.co/datasets/tiiuae/reasoning-images/resolve/main/falcon-h1r-logo.png" alt="drawing" width="800"/>
16
+
17
+ # Falcon-H1R-7B-FP8
18
+
19
+ This repository presents post FP8 quantized **Falcon-H1R-7B-FP8** via NVIDIA Model Optimizer, enabling efficient inference while preserving the strong reasoning introduced in the paper [Falcon-H1R: Pushing the Reasoning Frontiers with a Hybrid Model for Efficient Test-Time Scaling](https://huggingface.co/papers/2601.02346).
20
+
21
+ [Falcon-H1R-7B](https://huggingface.co/tiiuae/Falcon-H1R-7B) was trained via cold-start supervised fine-tuning with long reasoning traces and further enhanced by scaling RL with GRPO. The model demonstrates outstanding performance across various benchmark evaluations, including mathematics, programming, instruction following, and general logic.
22
+
23
+ ## Model Description
24
+
25
+ - **Developed by:** [Technology Innovation Institute](https://www.tii.ae)
26
+ - **Model type:** Causal decoder-only
27
+ - **Architecture:** Hybrid (Transformers + Mamba2) architecture
28
+ - **Language(s):** English, Multilingual
29
+ - **License:** [Falcon-LLM License](https://falconllm.tii.ae/falcon-terms-and-conditions.html)
30
+
31
+ ## Details
32
+
33
+ For more details on FP8 post-quantization for this model, please refer to the [Falcon-H1R-FP8 technical blogpost](https://falcon-lm.github.io/blog/falcon-h1r-7b-fp8).
34
+
35
+ For more details about the training protocol of this model, please refer to the [Falcon-H1R technical blogpost](https://falcon-lm.github.io/blog/falcon-h1r-7b) and [Technical Report](https://github.com/tiiuae/falcon-h1r/blob/main/tech_report.pdf).
36
+
37
+ # Usage
38
+
39
+ Currently to use this model, you can either rely on Hugging Face `transformers`, `vLLM` or `SGLang` library.
40
+
41
+ ## Inference
42
+
43
+ Make sure to install the latest version of `transformers` or `vLLM` or `SGLang`.
44
+
45
+ ```bash
46
+ pip install transformers
47
+ pip install mamba-ssm[causal-conv1d]
48
+ ```
49
+
50
+ For vLLM, make sure to install latest `vllm`:
51
+
52
+ ```bash
53
+ pip install vllm --extra-index-url https://wheels.vllm.ai/nightly
54
+ ```
55
+
56
+ - **FP8 support (vLLM)**: FP8 enablement from vLLM PR #32728 has been merged: [PR](https://github.com/vllm-project/vllm/pull/32728).
57
+ - **Memory footprint**: Model weight memory drops from 14.2 GB (BF16) to 7.9 GB (FP8).
58
+ - **Throughput**: Inference throughput improves from ~1.2× to up to 1.5×, depending on batch size, prompt length, and generation length, with minimal accuracy impact for this post-training FP8-quantized model.
59
+ - **Tensor parallelism (quantized models)**: Tensor parallel enablement from vLLM PR #33257 has been merged: [PR](https://github.com/vllm-project/vllm/pull/33257).
60
+
61
+ ## Sampling Parameters
62
+
63
+ We recommend using a **temperature** of **0.6** and **top-p** as **0.95** with max new tokens up to 65536.
64
+ For supported frameworks, you can adjust the repetition_penalty and presence_penalty parameters to reduce endless repetitions.
65
+
66
+ ## vLLM
67
+
68
+ For vLLM, simply start a server by executing the command below:
69
+
70
+ <details>
71
+ <summary> Click to expand </summary>
72
+ <br>
73
+
74
+ ```python
75
+ vllm serve tiiuae/Falcon-H1R-7B-FP8 \
76
+ --tensor-parallel-size 1 \
77
+ --data-parallel-size 1 \
78
+ --reasoning-parser deepseek_r1
79
+ --quantization modelopt
80
+ ```
81
+ <br>
82
+ Additional flags:
83
+
84
+ * You can reduce `--max-model-len` to preserve memory. Default value is `262144` which is quite large but not necessary for most scenarios.
85
+ * For function calling, append `--enable-auto-tool-choice` and `--tool-call-parser hermes` to the vllm serve command.
86
+
87
+
88
+ vLLM client execution code:
89
+ ```python
90
+ from openai import OpenAI
91
+ import json
92
+
93
+ client = OpenAI(
94
+ base_url="http://localhost:8000/v1",
95
+ api_key="EMPTY",
96
+ )
97
+
98
+ completion = client.chat.completions.create(
99
+ model="tiiuae/Falcon-H1R-7B-FP8",
100
+ messages=[
101
+ {"role": "user", "content": "If the product of two numbers is 360 and their GCD is 6, what is their LCM?"},
102
+ ],
103
+ temperature=0.6,
104
+ top_p=0.95,
105
+ max_tokens=65536
106
+ )
107
+
108
+ msg = completion.choices[0].message
109
+
110
+ print(json.dumps({
111
+ "reasoning": msg.reasoning_content,
112
+ "answer": msg.content
113
+ }, indent=2))
114
+ ```
115
+ </details>
116
+ <br>
117
+
118
+ # Evaluation
119
+
120
+ Falcon-H1R achieves state of art results in reasoning benchmarks. Please refer to the [Falcon-H1R-7B](https://huggingface.co/tiiuae/Falcon-H1R-7B) BF16 model card for full benchmark details.
121
+ The table below highlights accuracy comparisons on selected benchmarks between the Falcon-H1R BF16 checkpoint and the Falcon-H1R FP8 post-quantized checkpoint.
122
+
123
+ <table border="1" style="width: 100%; text-align: center; border-collapse: collapse;">
124
+ <colgroup>
125
+ <col style="width: 6%;">
126
+ <col style="width: 6%;">
127
+ <col style="width: 6%;">
128
+ </colgroup>
129
+ <thead>
130
+ <tr>
131
+ <th>Benchmark</th>
132
+ <th>Falcon-H1R-7B BF16</th>
133
+ <th>Falcon-H1R-7B FP8</th>
134
+ </tr>
135
+ </thead>
136
+ <tbody>
137
+ <tr>
138
+ <td>AIME25</td>
139
+ <td>83.1</td>
140
+ <td>82.3</td>
141
+ </tr>
142
+ <tr>
143
+ <td>LCBv5-v6</td>
144
+ <td>68.6</td>
145
+ <td>67.6</td>
146
+ </tr>
147
+ <tr>
148
+ <td>GPQA-D</td>
149
+ <td>61.3</td>
150
+ <td>61.2</td>
151
+ </tr>
152
+ </tbody>
153
+ </table>
154
+
155
+ # Useful links
156
+
157
+ - View [our FP8 release blogpost](https://falcon-lm.github.io/blog/falcon-h1r-7b-fp8).
158
+ - View [our release blogpost](https://falcon-lm.github.io/blog/falcon-h1r-7b).
159
+ - View [our technical report](https://huggingface.co/papers/2601.02346).
160
+ - Feel free to join [our discord server](https://discord.gg/Cbek57PrZE) if you have any questions or to interact with our researchers and developers.
161
+
162
+ # Acknowledgements
163
+ We sincerely thank the NVIDIA team — Sergio Perez, Shengliang Xu, Vadim Gimpelson, Mireille Fares, Liana Mikaelyan, Amit Kushwaha, and Adam Czekalowski — for their valuable collaboration and support in post-quantizing [Falcon-H1R-7B](https://huggingface.co/tiiuae/Falcon-H1R-7B) to FP8.
164
+
165
+ # Citation
166
+
167
+ If the Falcon-H1R family of reasoning models is helpful to your work, feel free to give us a cite.
168
+
169
+ ```
170
+ @misc{falcon-h1r,
171
+ title={Falcon-H1R: Pushing the Reasoning Frontiers with a Hybrid Model for Efficient Test-Time Scaling},
172
+ author={Falcon LLM Team and Iheb Chaabane and Puneesh Khanna and Suhail Mohmad and Slim Frikha and Shi Hu and Abdalgader Abubaker and Reda Alami and Mikhail Lubinets and Mohamed El Amine Seddik and Hakim Hacid},
173
+ year={2026},
174
+ eprint={2601.02346},
175
+ archivePrefix={arXiv},
176
+ primaryClass={cs.AI},
177
+ url={https://arxiv.org/abs/2601.02346},
178
+ }
179
+ ```
chat_template.jinja ADDED
@@ -0,0 +1,101 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {# --- System Prompt Handling --- #}
2
+ {%- if messages and messages[0]['role'] == 'system' %}
3
+ {% set system_msg = messages[0]['content'] %}
4
+ {%- set remaining_messages = messages[1:] %}
5
+ {%- else %}
6
+ {% set system_msg = "You are Falcon, a helpful AI assistant created by Technology Innovation Institute (TII). To answer the user's question, you first think about the reasoning process and then provide the user with the answer. The reasoning process is enclosed within <think> </think> tags, i.e., <think> reasoning process here </think> answer here." %}
7
+ {%- set remaining_messages = messages %}
8
+ {%- endif %}
9
+
10
+ {%- if tools %}
11
+ <|im_start|>system
12
+ {{ system_msg }}
13
+ # Tools
14
+ You may call one or more functions to assist with the user query. You are provided with function signatures within <tools></tools> XML tags.
15
+ <tools>
16
+ {%- for tool in tools %}
17
+ {{- "" }}
18
+ {{ tool | tojson }}
19
+ {%- endfor %}
20
+ {{- "" }}
21
+ </tools>
22
+ For each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:
23
+ <tool_call>
24
+ {"name": <function-name>, "arguments": <args-json-object>}
25
+ </tool_call>
26
+ <|im_end|>
27
+
28
+ {%- else %}
29
+ <|im_start|>system
30
+ {{ system_msg }}
31
+ <|im_end|>
32
+ {%- endif %}
33
+
34
+ {# --- Render remaining messages --- #}
35
+ {%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
36
+ {%- for message in messages[::-1] %}
37
+ {%- set index = (messages|length - 1) - loop.index0 %}
38
+ {%- if ns.multi_step_tool and message.role == "user" and message.content is string and not(message.content.startswith('<tool_response>') and message.content.endswith('</tool_response>')) %}
39
+ {%- set ns.multi_step_tool = false %}
40
+ {%- set ns.last_query_index = index %}
41
+ {%- endif %}
42
+ {%- endfor %}{%- for message in remaining_messages %}
43
+ {%- set content = message.get('content','') %}
44
+ {%- if message['role'] == 'user' %}
45
+ {{- '<|im_start|>' + message['role'] + '\n' + content + '<|im_end|>\n' }}
46
+ {%- elif message['role'] == 'assistant' %}
47
+ {{- '<|im_start|>' + message.role + '\n' }}
48
+ {%- set reasoning_content = '' %}
49
+ {%- if message.reasoning_content is string %}
50
+ {%- set reasoning_content = message.reasoning_content %}
51
+ {%- else %}
52
+ {%- if '</think>' in content %}
53
+ {%- set reasoning_content = content.split('</think>')[0].rstrip('\n').split('<think>')[-1].lstrip('\n') %}
54
+ {%- set content = content.split('</think>')[-1].lstrip('\n') %}
55
+ {%- endif %}
56
+ {%- endif %}
57
+ {%- if loop.index0 > ns.last_query_index %}
58
+ {%- if loop.last or (not loop.last and reasoning_content) %}
59
+ {{- '<think>\n' + reasoning_content.strip('\n') + '\n</think>\n\n' + content.lstrip('\n') }}
60
+ {%- else %}
61
+ {{- content + '\n' }}
62
+ {%- endif %}
63
+ {%- else %}
64
+ {{- content + '\n' }}
65
+ {%- endif %}
66
+ {%- if tools and message.tool_calls %}
67
+ {%- for tool_call in message.tool_calls %}
68
+ {%- if tool_call.function is defined %}
69
+ {%- set tool_call = tool_call.function %}
70
+ {%- endif %}
71
+ {{-'<tool_call>\n' }}
72
+ {{- '{"name": "'+ tool_call.name + '", "arguments":' }}
73
+ {%- if tool_call.arguments is string -%}
74
+ {{ tool_call.arguments }}
75
+ {%- else -%}
76
+ {{ tool_call.arguments | tojson }}
77
+ {%- endif -%}
78
+ {{- '}' }}
79
+ {{- '\n</tool_call>\n' }}
80
+ {%- endfor %}
81
+ {%- endif %}
82
+ {%- if not loop.last %}
83
+ {{- '<|im_end|>' + '\n' }}
84
+ {%- else %}
85
+ {{- '<|im_end|>' }}
86
+ {%- endif %}
87
+ {%- elif message['role'] == 'tool' %}
88
+ {# Tool responses treated as user messages #}
89
+ {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != "tool") %}
90
+ {{- '<|im_start|>user' }}
91
+ {%- endif %}
92
+ {{- '\n<tool_response>\n' + message['content'] + '\n</tool_response>' }}
93
+ {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
94
+ {{- '<|im_end|>\n' }}
95
+ {%- endif %}
96
+ {%- endif %}
97
+ {# --- Add generation prompt after last message if requested --- #}
98
+ {%- if loop.last and add_generation_prompt %}
99
+ {{- '<|im_start|>assistant\n' }}
100
+ {%- endif %}
101
+ {%- endfor %}
config.json ADDED
@@ -0,0 +1,97 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "FalconH1ForCausalLM"
4
+ ],
5
+ "attention_bias": false,
6
+ "attention_dropout": 0.0,
7
+ "attention_in_multiplier": 1.0,
8
+ "attention_out_multiplier": 0.10416666666666669,
9
+ "attn_layer_indices": null,
10
+ "bos_token_id": 1,
11
+ "dtype": "bfloat16",
12
+ "embedding_multiplier": 5.656854249492381,
13
+ "eos_token_id": 11,
14
+ "head_dim": 128,
15
+ "hidden_act": "silu",
16
+ "hidden_size": 3072,
17
+ "initializer_range": 0.02,
18
+ "intermediate_size": 12288,
19
+ "key_multiplier": 0.030690398488999456,
20
+ "lm_head_multiplier": 0.013020833333333334,
21
+ "mamba_chunk_size": 128,
22
+ "mamba_conv_bias": true,
23
+ "mamba_d_conv": 4,
24
+ "mamba_d_head": 128,
25
+ "mamba_d_ssm": 3072,
26
+ "mamba_d_state": 256,
27
+ "mamba_expand": 2,
28
+ "mamba_n_groups": 1,
29
+ "mamba_n_heads": 24,
30
+ "mamba_norm_before_gate": false,
31
+ "mamba_proj_bias": false,
32
+ "mamba_rms_norm": true,
33
+ "mamba_use_mlp": true,
34
+ "max_position_embeddings": 262144,
35
+ "mlp_bias": false,
36
+ "mlp_expansion_factor": 8,
37
+ "mlp_multipliers": [
38
+ 0.2946278254943948,
39
+ 0.032552083333333336
40
+ ],
41
+ "model_type": "falcon_h1",
42
+ "num_attention_heads": 12,
43
+ "num_hidden_layers": 44,
44
+ "num_key_value_heads": 2,
45
+ "num_logits_to_keep": 1,
46
+ "pad_token_id": 0,
47
+ "projectors_bias": false,
48
+ "rms_norm_eps": 1e-05,
49
+ "rope_scaling": null,
50
+ "rope_theta": 100000000000,
51
+ "ssm_in_multiplier": 0.4166666666666667,
52
+ "ssm_multipliers": [
53
+ 0.3535533905932738,
54
+ 0.25,
55
+ 0.1767766952966369,
56
+ 0.5,
57
+ 0.3535533905932738
58
+ ],
59
+ "ssm_out_multiplier": 0.11785113019775792,
60
+ "tie_word_embeddings": false,
61
+ "transformers_version": "4.57.0",
62
+ "use_cache": true,
63
+ "vocab_size": 130048,
64
+ "quantization_config": {
65
+ "config_groups": {
66
+ "group_0": {
67
+ "input_activations": {
68
+ "dynamic": false,
69
+ "num_bits": 8,
70
+ "type": "float"
71
+ },
72
+ "weights": {
73
+ "dynamic": false,
74
+ "num_bits": 8,
75
+ "type": "float"
76
+ },
77
+ "targets": [
78
+ "Linear"
79
+ ]
80
+ }
81
+ },
82
+ "ignore": [
83
+ "lm_head"
84
+ ],
85
+ "quant_algo": "FP8",
86
+ "kv_cache_scheme": {
87
+ "dynamic": false,
88
+ "num_bits": 8,
89
+ "type": "float"
90
+ },
91
+ "producer": {
92
+ "name": "modelopt",
93
+ "version": "0.41.0rc2.dev48+g5cc2a5451"
94
+ },
95
+ "quant_method": "modelopt"
96
+ }
97
+ }
generation_config.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 1,
4
+ "eos_token_id": [11, 228],
5
+ "pad_token_id": 0,
6
+ "transformers_version": "4.52.0.dev0"
7
+ }
hf_quant_config.json ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "producer": {
3
+ "name": "modelopt",
4
+ "version": "0.41.0rc2.dev48+g5cc2a5451"
5
+ },
6
+ "quantization": {
7
+ "quant_algo": "FP8",
8
+ "kv_cache_quant_algo": "FP8",
9
+ "exclude_modules": [
10
+ "lm_head"
11
+ ]
12
+ }
13
+ }
model-00001-of-00002.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4d54e9a447e1a1a7a3c9a31c2bb03310c2134e109cc456d64a20b4f859159155
3
+ size 4964381624
model-00002-of-00002.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3f38f03d0de5a97b1138ce2d91aeb9661f2a39bfa4a2517e299cea0150dca35b
3
+ size 3421659520
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 +1,856 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "additional_special_tokens": [
3
+ "<|pad|>",
4
+ ">>ABSTRACT<<",
5
+ ">>INTRODUCTION<<",
6
+ ">>SUMMARY<<",
7
+ ">>COMMENT<<",
8
+ ">>ANSWER<<",
9
+ ">>QUESTION<<",
10
+ ">>DOMAIN<<",
11
+ ">>PREFIX<<",
12
+ ">>SUFFIX<<",
13
+ ">>MIDDLE<<",
14
+ "<|finetune_right_pad_id|>",
15
+ "<|start_header_id|>",
16
+ "<|end_header_id|>",
17
+ "<|eom_id|>",
18
+ "<|eot_id|>",
19
+ "<|begin_of_text|>",
20
+ ">>TITLE<<",
21
+ "<schema>",
22
+ "</schema>",
23
+ "<scratch_pad>",
24
+ "</scratch_pad>",
25
+ "<explanation>",
26
+ "</explanation>",
27
+ "<file_sep>",
28
+ "<repo_name>",
29
+ "<|im_end|>",
30
+ "<|im_start|>",
31
+ ">>UNUSED_119<<",
32
+ ">>UNUSED_120<<",
33
+ "<|system|>",
34
+ ">>UNUSED_218<<",
35
+ ">>UNUSED_219<<",
36
+ ">>UNUSED_220<<",
37
+ ">>UNUSED_221<<",
38
+ ">>UNUSED_222<<",
39
+ ">>UNUSED_223<<",
40
+ ">>UNUSED_224<<",
41
+ ">>UNUSED_225<<",
42
+ ">>UNUSED_226<<",
43
+ ">>UNUSED_227<<",
44
+ ">>UNUSED_228<<",
45
+ ">>UNUSED_229<<",
46
+ ">>UNUSED_230<<",
47
+ ">>UNUSED_231<<",
48
+ ">>UNUSED_232<<",
49
+ ">>UNUSED_233<<",
50
+ ">>UNUSED_234<<",
51
+ ">>UNUSED_235<<",
52
+ ">>UNUSED_236<<",
53
+ ">>UNUSED_237<<",
54
+ ">>UNUSED_238<<",
55
+ ">>UNUSED_239<<",
56
+ ">>UNUSED_240<<",
57
+ ">>UNUSED_241<<",
58
+ ">>UNUSED_242<<",
59
+ ">>UNUSED_243<<",
60
+ ">>UNUSED_244<<",
61
+ ">>UNUSED_245<<",
62
+ ">>UNUSED_246<<",
63
+ ">>UNUSED_247<<",
64
+ ">>UNUSED_248<<",
65
+ ">>UNUSED_249<<",
66
+ ">>UNUSED_250<<",
67
+ ">>UNUSED_251<<",
68
+ ">>UNUSED_252<<",
69
+ ">>UNUSED_253<<",
70
+ ">>UNUSED_254<<",
71
+ ">>UNUSED_255<<",
72
+ ">>UNUSED_256<<",
73
+ ">>UNUSED_257<<",
74
+ ">>UNUSED_258<<",
75
+ ">>UNUSED_259<<",
76
+ ">>UNUSED_260<<",
77
+ ">>UNUSED_261<<",
78
+ ">>UNUSED_262<<",
79
+ ">>UNUSED_263<<",
80
+ ">>UNUSED_264<<",
81
+ ">>UNUSED_265<<",
82
+ ">>UNUSED_266<<",
83
+ ">>UNUSED_267<<",
84
+ ">>UNUSED_268<<",
85
+ ">>UNUSED_269<<",
86
+ ">>UNUSED_270<<",
87
+ ">>UNUSED_271<<",
88
+ ">>UNUSED_272<<",
89
+ ">>UNUSED_273<<",
90
+ ">>UNUSED_274<<",
91
+ ">>UNUSED_275<<",
92
+ ">>UNUSED_276<<",
93
+ ">>UNUSED_277<<",
94
+ ">>UNUSED_278<<",
95
+ ">>UNUSED_279<<",
96
+ ">>UNUSED_280<<",
97
+ ">>UNUSED_281<<",
98
+ ">>UNUSED_282<<",
99
+ ">>UNUSED_283<<",
100
+ ">>UNUSED_284<<",
101
+ ">>UNUSED_285<<",
102
+ ">>UNUSED_286<<",
103
+ ">>UNUSED_287<<",
104
+ ">>UNUSED_288<<",
105
+ ">>UNUSED_289<<",
106
+ ">>UNUSED_290<<",
107
+ ">>UNUSED_291<<",
108
+ ">>UNUSED_292<<",
109
+ ">>UNUSED_293<<",
110
+ ">>UNUSED_294<<",
111
+ ">>UNUSED_295<<",
112
+ ">>UNUSED_296<<",
113
+ ">>UNUSED_297<<",
114
+ ">>UNUSED_298<<",
115
+ ">>UNUSED_299<<",
116
+ ">>UNUSED_300<<",
117
+ ">>UNUSED_301<<",
118
+ ">>UNUSED_302<<",
119
+ ">>UNUSED_303<<",
120
+ ">>UNUSED_304<<",
121
+ ">>UNUSED_305<<",
122
+ ">>UNUSED_306<<",
123
+ ">>UNUSED_307<<",
124
+ ">>UNUSED_308<<",
125
+ ">>UNUSED_309<<",
126
+ ">>UNUSED_310<<",
127
+ ">>UNUSED_311<<",
128
+ ">>UNUSED_312<<",
129
+ ">>UNUSED_313<<",
130
+ ">>UNUSED_314<<",
131
+ ">>UNUSED_315<<",
132
+ ">>UNUSED_316<<",
133
+ ">>UNUSED_317<<",
134
+ ">>UNUSED_318<<",
135
+ ">>UNUSED_319<<",
136
+ ">>UNUSED_320<<",
137
+ ">>UNUSED_321<<",
138
+ ">>UNUSED_322<<",
139
+ ">>UNUSED_323<<",
140
+ ">>UNUSED_324<<",
141
+ ">>UNUSED_325<<",
142
+ ">>UNUSED_326<<",
143
+ ">>UNUSED_327<<",
144
+ ">>UNUSED_328<<",
145
+ ">>UNUSED_329<<",
146
+ ">>UNUSED_330<<",
147
+ ">>UNUSED_331<<",
148
+ ">>UNUSED_332<<",
149
+ ">>UNUSED_333<<",
150
+ ">>UNUSED_334<<",
151
+ ">>UNUSED_335<<",
152
+ ">>UNUSED_336<<",
153
+ ">>UNUSED_337<<",
154
+ ">>UNUSED_338<<",
155
+ ">>UNUSED_339<<",
156
+ ">>UNUSED_340<<",
157
+ ">>UNUSED_341<<",
158
+ ">>UNUSED_342<<",
159
+ ">>UNUSED_343<<",
160
+ ">>UNUSED_344<<",
161
+ ">>UNUSED_345<<",
162
+ ">>UNUSED_346<<",
163
+ ">>UNUSED_347<<",
164
+ ">>UNUSED_348<<",
165
+ ">>UNUSED_349<<",
166
+ ">>UNUSED_350<<",
167
+ ">>UNUSED_351<<",
168
+ ">>UNUSED_352<<",
169
+ ">>UNUSED_353<<",
170
+ ">>UNUSED_354<<",
171
+ ">>UNUSED_355<<",
172
+ ">>UNUSED_356<<",
173
+ ">>UNUSED_357<<",
174
+ ">>UNUSED_358<<",
175
+ ">>UNUSED_359<<",
176
+ ">>UNUSED_360<<",
177
+ ">>UNUSED_361<<",
178
+ ">>UNUSED_362<<",
179
+ ">>UNUSED_363<<",
180
+ ">>UNUSED_364<<",
181
+ ">>UNUSED_365<<",
182
+ ">>UNUSED_366<<",
183
+ ">>UNUSED_367<<",
184
+ ">>UNUSED_368<<",
185
+ ">>UNUSED_369<<",
186
+ ">>UNUSED_370<<",
187
+ ">>UNUSED_371<<",
188
+ ">>UNUSED_372<<",
189
+ ">>UNUSED_373<<",
190
+ ">>UNUSED_374<<",
191
+ ">>UNUSED_375<<",
192
+ ">>UNUSED_376<<",
193
+ ">>UNUSED_377<<",
194
+ ">>UNUSED_378<<",
195
+ ">>UNUSED_379<<",
196
+ ">>UNUSED_380<<",
197
+ ">>UNUSED_381<<",
198
+ ">>UNUSED_382<<",
199
+ ">>UNUSED_383<<",
200
+ ">>UNUSED_384<<",
201
+ ">>UNUSED_385<<",
202
+ ">>UNUSED_386<<",
203
+ ">>UNUSED_387<<",
204
+ ">>UNUSED_388<<",
205
+ ">>UNUSED_389<<",
206
+ ">>UNUSED_390<<",
207
+ ">>UNUSED_391<<",
208
+ ">>UNUSED_392<<",
209
+ ">>UNUSED_393<<",
210
+ ">>UNUSED_394<<",
211
+ ">>UNUSED_395<<",
212
+ ">>UNUSED_396<<",
213
+ ">>UNUSED_397<<",
214
+ ">>UNUSED_398<<",
215
+ ">>UNUSED_399<<",
216
+ ">>UNUSED_400<<",
217
+ ">>UNUSED_401<<",
218
+ ">>UNUSED_402<<",
219
+ ">>UNUSED_403<<",
220
+ ">>UNUSED_404<<",
221
+ ">>UNUSED_405<<",
222
+ ">>UNUSED_406<<",
223
+ ">>UNUSED_407<<",
224
+ ">>UNUSED_408<<",
225
+ ">>UNUSED_409<<",
226
+ ">>UNUSED_410<<",
227
+ ">>UNUSED_411<<",
228
+ ">>UNUSED_412<<",
229
+ ">>UNUSED_413<<",
230
+ ">>UNUSED_414<<",
231
+ ">>UNUSED_415<<",
232
+ ">>UNUSED_416<<",
233
+ ">>UNUSED_417<<",
234
+ ">>UNUSED_418<<",
235
+ ">>UNUSED_419<<",
236
+ ">>UNUSED_420<<",
237
+ ">>UNUSED_421<<",
238
+ ">>UNUSED_422<<",
239
+ ">>UNUSED_423<<",
240
+ ">>UNUSED_424<<",
241
+ ">>UNUSED_425<<",
242
+ ">>UNUSED_426<<",
243
+ ">>UNUSED_427<<",
244
+ ">>UNUSED_428<<",
245
+ ">>UNUSED_429<<",
246
+ ">>UNUSED_430<<",
247
+ ">>UNUSED_431<<",
248
+ ">>UNUSED_432<<",
249
+ ">>UNUSED_433<<",
250
+ ">>UNUSED_434<<",
251
+ ">>UNUSED_435<<",
252
+ ">>UNUSED_436<<",
253
+ ">>UNUSED_437<<",
254
+ ">>UNUSED_438<<",
255
+ ">>UNUSED_439<<",
256
+ ">>UNUSED_440<<",
257
+ ">>UNUSED_441<<",
258
+ ">>UNUSED_442<<",
259
+ ">>UNUSED_443<<",
260
+ ">>UNUSED_444<<",
261
+ ">>UNUSED_445<<",
262
+ ">>UNUSED_446<<",
263
+ ">>UNUSED_447<<",
264
+ ">>UNUSED_448<<",
265
+ ">>UNUSED_449<<",
266
+ ">>UNUSED_450<<",
267
+ ">>UNUSED_451<<",
268
+ ">>UNUSED_452<<",
269
+ ">>UNUSED_453<<",
270
+ ">>UNUSED_454<<",
271
+ ">>UNUSED_455<<",
272
+ ">>UNUSED_456<<",
273
+ ">>UNUSED_457<<",
274
+ ">>UNUSED_458<<",
275
+ ">>UNUSED_459<<",
276
+ ">>UNUSED_460<<",
277
+ ">>UNUSED_461<<",
278
+ ">>UNUSED_462<<",
279
+ ">>UNUSED_463<<",
280
+ ">>UNUSED_464<<",
281
+ ">>UNUSED_465<<",
282
+ ">>UNUSED_466<<",
283
+ ">>UNUSED_467<<",
284
+ ">>UNUSED_468<<",
285
+ ">>UNUSED_469<<",
286
+ ">>UNUSED_470<<",
287
+ ">>UNUSED_471<<",
288
+ ">>UNUSED_472<<",
289
+ ">>UNUSED_473<<",
290
+ ">>UNUSED_474<<",
291
+ ">>UNUSED_475<<",
292
+ ">>UNUSED_476<<",
293
+ ">>UNUSED_477<<",
294
+ ">>UNUSED_478<<",
295
+ ">>UNUSED_479<<",
296
+ ">>UNUSED_480<<",
297
+ ">>UNUSED_481<<",
298
+ ">>UNUSED_482<<",
299
+ ">>UNUSED_483<<",
300
+ ">>UNUSED_484<<",
301
+ ">>UNUSED_485<<",
302
+ ">>UNUSED_486<<",
303
+ ">>UNUSED_487<<",
304
+ ">>UNUSED_488<<",
305
+ ">>UNUSED_489<<",
306
+ ">>UNUSED_490<<",
307
+ ">>UNUSED_491<<",
308
+ ">>UNUSED_492<<",
309
+ ">>UNUSED_493<<",
310
+ ">>UNUSED_494<<",
311
+ ">>UNUSED_495<<",
312
+ ">>UNUSED_496<<",
313
+ ">>UNUSED_497<<",
314
+ ">>UNUSED_498<<",
315
+ ">>UNUSED_499<<",
316
+ ">>UNUSED_500<<",
317
+ ">>UNUSED_501<<",
318
+ ">>UNUSED_502<<",
319
+ ">>UNUSED_503<<",
320
+ ">>UNUSED_504<<",
321
+ ">>UNUSED_505<<",
322
+ ">>UNUSED_506<<",
323
+ ">>UNUSED_507<<",
324
+ ">>UNUSED_508<<",
325
+ ">>UNUSED_509<<",
326
+ ">>UNUSED_510<<",
327
+ ">>UNUSED_511<<",
328
+ ">>UNUSED_512<<",
329
+ ">>UNUSED_513<<",
330
+ ">>UNUSED_514<<",
331
+ ">>UNUSED_515<<",
332
+ ">>UNUSED_516<<",
333
+ ">>UNUSED_517<<",
334
+ ">>UNUSED_518<<",
335
+ ">>UNUSED_519<<",
336
+ ">>UNUSED_520<<",
337
+ ">>UNUSED_521<<",
338
+ ">>UNUSED_522<<",
339
+ ">>UNUSED_523<<",
340
+ ">>UNUSED_524<<",
341
+ ">>UNUSED_525<<",
342
+ ">>UNUSED_526<<",
343
+ ">>UNUSED_527<<",
344
+ ">>UNUSED_528<<",
345
+ ">>UNUSED_529<<",
346
+ ">>UNUSED_530<<",
347
+ ">>UNUSED_531<<",
348
+ ">>UNUSED_532<<",
349
+ ">>UNUSED_533<<",
350
+ ">>UNUSED_534<<",
351
+ ">>UNUSED_535<<",
352
+ ">>UNUSED_536<<",
353
+ ">>UNUSED_537<<",
354
+ ">>UNUSED_538<<",
355
+ ">>UNUSED_539<<",
356
+ ">>UNUSED_540<<",
357
+ ">>UNUSED_541<<",
358
+ ">>UNUSED_542<<",
359
+ ">>UNUSED_543<<",
360
+ ">>UNUSED_544<<",
361
+ ">>UNUSED_545<<",
362
+ ">>UNUSED_546<<",
363
+ ">>UNUSED_547<<",
364
+ ">>UNUSED_548<<",
365
+ ">>UNUSED_549<<",
366
+ ">>UNUSED_550<<",
367
+ ">>UNUSED_551<<",
368
+ ">>UNUSED_552<<",
369
+ ">>UNUSED_553<<",
370
+ ">>UNUSED_554<<",
371
+ ">>UNUSED_555<<",
372
+ ">>UNUSED_556<<",
373
+ ">>UNUSED_557<<",
374
+ ">>UNUSED_558<<",
375
+ ">>UNUSED_559<<",
376
+ ">>UNUSED_560<<",
377
+ ">>UNUSED_561<<",
378
+ ">>UNUSED_562<<",
379
+ ">>UNUSED_563<<",
380
+ ">>UNUSED_564<<",
381
+ ">>UNUSED_565<<",
382
+ ">>UNUSED_566<<",
383
+ ">>UNUSED_567<<",
384
+ ">>UNUSED_568<<",
385
+ ">>UNUSED_569<<",
386
+ ">>UNUSED_570<<",
387
+ ">>UNUSED_571<<",
388
+ ">>UNUSED_572<<",
389
+ ">>UNUSED_573<<",
390
+ ">>UNUSED_574<<",
391
+ ">>UNUSED_575<<",
392
+ ">>UNUSED_576<<",
393
+ ">>UNUSED_577<<",
394
+ ">>UNUSED_578<<",
395
+ ">>UNUSED_579<<",
396
+ ">>UNUSED_580<<",
397
+ ">>UNUSED_581<<",
398
+ ">>UNUSED_582<<",
399
+ ">>UNUSED_583<<",
400
+ ">>UNUSED_584<<",
401
+ ">>UNUSED_585<<",
402
+ ">>UNUSED_586<<",
403
+ ">>UNUSED_587<<",
404
+ ">>UNUSED_588<<",
405
+ ">>UNUSED_589<<",
406
+ ">>UNUSED_590<<",
407
+ ">>UNUSED_591<<",
408
+ ">>UNUSED_592<<",
409
+ ">>UNUSED_593<<",
410
+ ">>UNUSED_594<<",
411
+ ">>UNUSED_595<<",
412
+ ">>UNUSED_596<<",
413
+ ">>UNUSED_597<<",
414
+ ">>UNUSED_598<<",
415
+ ">>UNUSED_599<<",
416
+ ">>UNUSED_600<<",
417
+ ">>UNUSED_601<<",
418
+ ">>UNUSED_602<<",
419
+ ">>UNUSED_603<<",
420
+ ">>UNUSED_604<<",
421
+ ">>UNUSED_605<<",
422
+ ">>UNUSED_606<<",
423
+ ">>UNUSED_607<<",
424
+ ">>UNUSED_608<<",
425
+ ">>UNUSED_609<<",
426
+ ">>UNUSED_610<<",
427
+ ">>UNUSED_611<<",
428
+ ">>UNUSED_612<<",
429
+ ">>UNUSED_613<<",
430
+ ">>UNUSED_614<<",
431
+ ">>UNUSED_615<<",
432
+ ">>UNUSED_616<<",
433
+ ">>UNUSED_617<<",
434
+ ">>UNUSED_618<<",
435
+ ">>UNUSED_619<<",
436
+ ">>UNUSED_620<<",
437
+ ">>UNUSED_621<<",
438
+ ">>UNUSED_622<<",
439
+ ">>UNUSED_623<<",
440
+ ">>UNUSED_624<<",
441
+ ">>UNUSED_625<<",
442
+ ">>UNUSED_626<<",
443
+ ">>UNUSED_627<<",
444
+ ">>UNUSED_628<<",
445
+ ">>UNUSED_629<<",
446
+ ">>UNUSED_630<<",
447
+ ">>UNUSED_631<<",
448
+ ">>UNUSED_632<<",
449
+ ">>UNUSED_633<<",
450
+ ">>UNUSED_634<<",
451
+ ">>UNUSED_635<<",
452
+ ">>UNUSED_636<<",
453
+ ">>UNUSED_637<<",
454
+ ">>UNUSED_638<<",
455
+ ">>UNUSED_639<<",
456
+ ">>UNUSED_640<<",
457
+ ">>UNUSED_641<<",
458
+ ">>UNUSED_642<<",
459
+ ">>UNUSED_643<<",
460
+ ">>UNUSED_644<<",
461
+ ">>UNUSED_645<<",
462
+ ">>UNUSED_646<<",
463
+ ">>UNUSED_647<<",
464
+ ">>UNUSED_648<<",
465
+ ">>UNUSED_649<<",
466
+ ">>UNUSED_650<<",
467
+ ">>UNUSED_651<<",
468
+ ">>UNUSED_652<<",
469
+ ">>UNUSED_653<<",
470
+ ">>UNUSED_654<<",
471
+ ">>UNUSED_655<<",
472
+ ">>UNUSED_656<<",
473
+ ">>UNUSED_657<<",
474
+ ">>UNUSED_658<<",
475
+ ">>UNUSED_659<<",
476
+ ">>UNUSED_660<<",
477
+ ">>UNUSED_661<<",
478
+ ">>UNUSED_662<<",
479
+ ">>UNUSED_663<<",
480
+ ">>UNUSED_664<<",
481
+ ">>UNUSED_665<<",
482
+ ">>UNUSED_666<<",
483
+ ">>UNUSED_667<<",
484
+ ">>UNUSED_668<<",
485
+ ">>UNUSED_669<<",
486
+ ">>UNUSED_670<<",
487
+ ">>UNUSED_671<<",
488
+ ">>UNUSED_672<<",
489
+ ">>UNUSED_673<<",
490
+ ">>UNUSED_674<<",
491
+ ">>UNUSED_675<<",
492
+ ">>UNUSED_676<<",
493
+ ">>UNUSED_677<<",
494
+ ">>UNUSED_678<<",
495
+ ">>UNUSED_679<<",
496
+ ">>UNUSED_680<<",
497
+ ">>UNUSED_681<<",
498
+ ">>UNUSED_682<<",
499
+ ">>UNUSED_683<<",
500
+ ">>UNUSED_684<<",
501
+ ">>UNUSED_685<<",
502
+ ">>UNUSED_686<<",
503
+ ">>UNUSED_687<<",
504
+ ">>UNUSED_688<<",
505
+ ">>UNUSED_689<<",
506
+ ">>UNUSED_690<<",
507
+ ">>UNUSED_691<<",
508
+ ">>UNUSED_692<<",
509
+ ">>UNUSED_693<<",
510
+ ">>UNUSED_694<<",
511
+ ">>UNUSED_695<<",
512
+ ">>UNUSED_696<<",
513
+ ">>UNUSED_697<<",
514
+ ">>UNUSED_698<<",
515
+ ">>UNUSED_699<<",
516
+ ">>UNUSED_700<<",
517
+ ">>UNUSED_701<<",
518
+ ">>UNUSED_702<<",
519
+ ">>UNUSED_703<<",
520
+ ">>UNUSED_704<<",
521
+ ">>UNUSED_705<<",
522
+ ">>UNUSED_706<<",
523
+ ">>UNUSED_707<<",
524
+ ">>UNUSED_708<<",
525
+ ">>UNUSED_709<<",
526
+ ">>UNUSED_710<<",
527
+ ">>UNUSED_711<<",
528
+ ">>UNUSED_712<<",
529
+ ">>UNUSED_713<<",
530
+ ">>UNUSED_714<<",
531
+ ">>UNUSED_715<<",
532
+ ">>UNUSED_716<<",
533
+ ">>UNUSED_717<<",
534
+ ">>UNUSED_718<<",
535
+ ">>UNUSED_719<<",
536
+ ">>UNUSED_720<<",
537
+ ">>UNUSED_721<<",
538
+ ">>UNUSED_722<<",
539
+ ">>UNUSED_723<<",
540
+ ">>UNUSED_724<<",
541
+ ">>UNUSED_725<<",
542
+ ">>UNUSED_726<<",
543
+ ">>UNUSED_727<<",
544
+ ">>UNUSED_728<<",
545
+ ">>UNUSED_729<<",
546
+ ">>UNUSED_730<<",
547
+ ">>UNUSED_731<<",
548
+ ">>UNUSED_732<<",
549
+ ">>UNUSED_733<<",
550
+ ">>UNUSED_734<<",
551
+ ">>UNUSED_735<<",
552
+ ">>UNUSED_736<<",
553
+ ">>UNUSED_737<<",
554
+ ">>UNUSED_738<<",
555
+ ">>UNUSED_739<<",
556
+ ">>UNUSED_740<<",
557
+ ">>UNUSED_741<<",
558
+ ">>UNUSED_742<<",
559
+ ">>UNUSED_743<<",
560
+ ">>UNUSED_744<<",
561
+ ">>UNUSED_745<<",
562
+ ">>UNUSED_746<<",
563
+ ">>UNUSED_747<<",
564
+ ">>UNUSED_748<<",
565
+ ">>UNUSED_749<<",
566
+ ">>UNUSED_750<<",
567
+ ">>UNUSED_751<<",
568
+ ">>UNUSED_752<<",
569
+ ">>UNUSED_753<<",
570
+ ">>UNUSED_754<<",
571
+ ">>UNUSED_755<<",
572
+ ">>UNUSED_756<<",
573
+ ">>UNUSED_757<<",
574
+ ">>UNUSED_758<<",
575
+ ">>UNUSED_759<<",
576
+ ">>UNUSED_760<<",
577
+ ">>UNUSED_761<<",
578
+ ">>UNUSED_762<<",
579
+ ">>UNUSED_763<<",
580
+ ">>UNUSED_764<<",
581
+ ">>UNUSED_765<<",
582
+ ">>UNUSED_766<<",
583
+ ">>UNUSED_767<<",
584
+ ">>UNUSED_768<<",
585
+ ">>UNUSED_769<<",
586
+ ">>UNUSED_770<<",
587
+ ">>UNUSED_771<<",
588
+ ">>UNUSED_772<<",
589
+ ">>UNUSED_773<<",
590
+ ">>UNUSED_774<<",
591
+ ">>UNUSED_775<<",
592
+ ">>UNUSED_776<<",
593
+ ">>UNUSED_777<<",
594
+ ">>UNUSED_778<<",
595
+ ">>UNUSED_779<<",
596
+ ">>UNUSED_780<<",
597
+ ">>UNUSED_781<<",
598
+ ">>UNUSED_782<<",
599
+ ">>UNUSED_783<<",
600
+ ">>UNUSED_784<<",
601
+ ">>UNUSED_785<<",
602
+ ">>UNUSED_786<<",
603
+ ">>UNUSED_787<<",
604
+ ">>UNUSED_788<<",
605
+ ">>UNUSED_789<<",
606
+ ">>UNUSED_790<<",
607
+ ">>UNUSED_791<<",
608
+ ">>UNUSED_792<<",
609
+ ">>UNUSED_793<<",
610
+ ">>UNUSED_794<<",
611
+ ">>UNUSED_795<<",
612
+ ">>UNUSED_796<<",
613
+ ">>UNUSED_797<<",
614
+ ">>UNUSED_798<<",
615
+ ">>UNUSED_799<<",
616
+ ">>UNUSED_800<<",
617
+ ">>UNUSED_801<<",
618
+ ">>UNUSED_802<<",
619
+ ">>UNUSED_803<<",
620
+ ">>UNUSED_804<<",
621
+ ">>UNUSED_805<<",
622
+ ">>UNUSED_806<<",
623
+ ">>UNUSED_807<<",
624
+ ">>UNUSED_808<<",
625
+ ">>UNUSED_809<<",
626
+ ">>UNUSED_810<<",
627
+ ">>UNUSED_811<<",
628
+ ">>UNUSED_812<<",
629
+ ">>UNUSED_813<<",
630
+ ">>UNUSED_814<<",
631
+ ">>UNUSED_815<<",
632
+ ">>UNUSED_816<<",
633
+ ">>UNUSED_817<<",
634
+ ">>UNUSED_818<<",
635
+ ">>UNUSED_819<<",
636
+ ">>UNUSED_820<<",
637
+ ">>UNUSED_821<<",
638
+ ">>UNUSED_822<<",
639
+ ">>UNUSED_823<<",
640
+ ">>UNUSED_824<<",
641
+ ">>UNUSED_825<<",
642
+ ">>UNUSED_826<<",
643
+ ">>UNUSED_827<<",
644
+ ">>UNUSED_828<<",
645
+ ">>UNUSED_829<<",
646
+ ">>UNUSED_830<<",
647
+ ">>UNUSED_831<<",
648
+ ">>UNUSED_832<<",
649
+ ">>UNUSED_833<<",
650
+ ">>UNUSED_834<<",
651
+ ">>UNUSED_835<<",
652
+ ">>UNUSED_836<<",
653
+ ">>UNUSED_837<<",
654
+ ">>UNUSED_838<<",
655
+ ">>UNUSED_839<<",
656
+ ">>UNUSED_840<<",
657
+ ">>UNUSED_841<<",
658
+ ">>UNUSED_842<<",
659
+ ">>UNUSED_843<<",
660
+ ">>UNUSED_844<<",
661
+ ">>UNUSED_845<<",
662
+ ">>UNUSED_846<<",
663
+ ">>UNUSED_847<<",
664
+ ">>UNUSED_848<<",
665
+ ">>UNUSED_849<<",
666
+ ">>UNUSED_850<<",
667
+ ">>UNUSED_851<<",
668
+ ">>UNUSED_852<<",
669
+ ">>UNUSED_853<<",
670
+ ">>UNUSED_854<<",
671
+ ">>UNUSED_855<<",
672
+ ">>UNUSED_856<<",
673
+ ">>UNUSED_857<<",
674
+ ">>UNUSED_858<<",
675
+ ">>UNUSED_859<<",
676
+ ">>UNUSED_860<<",
677
+ ">>UNUSED_861<<",
678
+ ">>UNUSED_862<<",
679
+ ">>UNUSED_863<<",
680
+ ">>UNUSED_864<<",
681
+ ">>UNUSED_865<<",
682
+ ">>UNUSED_866<<",
683
+ ">>UNUSED_867<<",
684
+ ">>UNUSED_868<<",
685
+ ">>UNUSED_869<<",
686
+ ">>UNUSED_870<<",
687
+ ">>UNUSED_871<<",
688
+ ">>UNUSED_872<<",
689
+ ">>UNUSED_873<<",
690
+ ">>UNUSED_874<<",
691
+ ">>UNUSED_875<<",
692
+ ">>UNUSED_876<<",
693
+ ">>UNUSED_877<<",
694
+ ">>UNUSED_878<<",
695
+ ">>UNUSED_879<<",
696
+ ">>UNUSED_880<<",
697
+ ">>UNUSED_881<<",
698
+ ">>UNUSED_882<<",
699
+ ">>UNUSED_883<<",
700
+ ">>UNUSED_884<<",
701
+ ">>UNUSED_885<<",
702
+ ">>UNUSED_886<<",
703
+ ">>UNUSED_887<<",
704
+ ">>UNUSED_888<<",
705
+ ">>UNUSED_889<<",
706
+ ">>UNUSED_890<<",
707
+ ">>UNUSED_891<<",
708
+ ">>UNUSED_892<<",
709
+ ">>UNUSED_893<<",
710
+ ">>UNUSED_894<<",
711
+ ">>UNUSED_895<<",
712
+ ">>UNUSED_896<<",
713
+ ">>UNUSED_897<<",
714
+ ">>UNUSED_898<<",
715
+ ">>UNUSED_899<<",
716
+ ">>UNUSED_900<<",
717
+ ">>UNUSED_901<<",
718
+ ">>UNUSED_902<<",
719
+ ">>UNUSED_903<<",
720
+ ">>UNUSED_904<<",
721
+ ">>UNUSED_905<<",
722
+ ">>UNUSED_906<<",
723
+ ">>UNUSED_907<<",
724
+ ">>UNUSED_908<<",
725
+ ">>UNUSED_909<<",
726
+ ">>UNUSED_910<<",
727
+ ">>UNUSED_911<<",
728
+ ">>UNUSED_912<<",
729
+ ">>UNUSED_913<<",
730
+ ">>UNUSED_914<<",
731
+ ">>UNUSED_915<<",
732
+ ">>UNUSED_916<<",
733
+ ">>UNUSED_917<<",
734
+ ">>UNUSED_918<<",
735
+ ">>UNUSED_919<<",
736
+ ">>UNUSED_920<<",
737
+ ">>UNUSED_921<<",
738
+ ">>UNUSED_922<<",
739
+ ">>UNUSED_923<<",
740
+ ">>UNUSED_924<<",
741
+ ">>UNUSED_925<<",
742
+ ">>UNUSED_926<<",
743
+ ">>UNUSED_927<<",
744
+ ">>UNUSED_928<<",
745
+ ">>UNUSED_929<<",
746
+ ">>UNUSED_930<<",
747
+ ">>UNUSED_931<<",
748
+ ">>UNUSED_932<<",
749
+ ">>UNUSED_933<<",
750
+ ">>UNUSED_934<<",
751
+ ">>UNUSED_935<<",
752
+ ">>UNUSED_936<<",
753
+ ">>UNUSED_937<<",
754
+ ">>UNUSED_938<<",
755
+ ">>UNUSED_939<<",
756
+ ">>UNUSED_940<<",
757
+ ">>UNUSED_941<<",
758
+ ">>UNUSED_942<<",
759
+ ">>UNUSED_943<<",
760
+ ">>UNUSED_944<<",
761
+ ">>UNUSED_945<<",
762
+ ">>UNUSED_946<<",
763
+ ">>UNUSED_947<<",
764
+ ">>UNUSED_948<<",
765
+ ">>UNUSED_949<<",
766
+ ">>UNUSED_950<<",
767
+ ">>UNUSED_951<<",
768
+ ">>UNUSED_952<<",
769
+ ">>UNUSED_953<<",
770
+ ">>UNUSED_954<<",
771
+ ">>UNUSED_955<<",
772
+ ">>UNUSED_956<<",
773
+ ">>UNUSED_957<<",
774
+ ">>UNUSED_958<<",
775
+ ">>UNUSED_959<<",
776
+ ">>UNUSED_960<<",
777
+ ">>UNUSED_961<<",
778
+ ">>UNUSED_962<<",
779
+ ">>UNUSED_963<<",
780
+ ">>UNUSED_964<<",
781
+ ">>UNUSED_965<<",
782
+ ">>UNUSED_966<<",
783
+ ">>UNUSED_967<<",
784
+ ">>UNUSED_968<<",
785
+ ">>UNUSED_969<<",
786
+ ">>UNUSED_970<<",
787
+ ">>UNUSED_971<<",
788
+ ">>UNUSED_972<<",
789
+ ">>UNUSED_973<<",
790
+ ">>UNUSED_974<<",
791
+ ">>UNUSED_975<<",
792
+ ">>UNUSED_976<<",
793
+ ">>UNUSED_977<<",
794
+ ">>UNUSED_978<<",
795
+ ">>UNUSED_979<<",
796
+ ">>UNUSED_980<<",
797
+ ">>UNUSED_981<<",
798
+ ">>UNUSED_982<<",
799
+ ">>UNUSED_983<<",
800
+ ">>UNUSED_984<<",
801
+ ">>UNUSED_985<<",
802
+ ">>UNUSED_986<<",
803
+ ">>UNUSED_987<<",
804
+ ">>UNUSED_988<<",
805
+ ">>UNUSED_989<<",
806
+ ">>UNUSED_990<<",
807
+ ">>UNUSED_991<<",
808
+ ">>UNUSED_992<<",
809
+ ">>UNUSED_993<<",
810
+ ">>UNUSED_994<<",
811
+ ">>UNUSED_995<<",
812
+ ">>UNUSED_996<<",
813
+ ">>UNUSED_997<<",
814
+ ">>UNUSED_998<<",
815
+ ">>UNUSED_999<<",
816
+ ">>UNUSED_1000<<",
817
+ ">>UNUSED_1001<<",
818
+ ">>UNUSED_1002<<",
819
+ ">>UNUSED_1003<<",
820
+ ">>UNUSED_1004<<",
821
+ ">>UNUSED_1005<<",
822
+ ">>UNUSED_1006<<",
823
+ ">>UNUSED_1007<<",
824
+ ">>UNUSED_1008<<",
825
+ ">>UNUSED_1009<<",
826
+ ">>UNUSED_1010<<",
827
+ ">>UNUSED_1011<<",
828
+ ">>UNUSED_1012<<",
829
+ ">>UNUSED_1013<<",
830
+ ">>UNUSED_1014<<",
831
+ ">>UNUSED_1015<<",
832
+ ">>UNUSED_1016<<",
833
+ ">>UNUSED_1017<<",
834
+ ">>UNUSED_1018<<",
835
+ ">>UNUSED_1019<<",
836
+ ">>UNUSED_1020<<",
837
+ ">>UNUSED_1021<<",
838
+ ">>UNUSED_1022<<",
839
+ ">>UNUSED_1023<<"
840
+ ],
841
+ "bos_token": {
842
+ "content": "<|begin_of_text|>",
843
+ "lstrip": false,
844
+ "normalized": false,
845
+ "rstrip": false,
846
+ "single_word": false
847
+ },
848
+ "eos_token": {
849
+ "content": "<|end_of_text|>",
850
+ "lstrip": false,
851
+ "normalized": false,
852
+ "rstrip": false,
853
+ "single_word": false
854
+ },
855
+ "pad_token": "<|end_of_text|>"
856
+ }
tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a1a906251e685c2ff1fbb0ab2e36d9d1b276e6283a7a9a3e721c5b1923c017c0
3
+ size 10534300
tokenizer_config.json ADDED
The diff for this file is too large to render. See raw diff