Hugging Face's logo

xammi
/
MunicipalPredictionModel-TinyLlama

PEFT
TensorBoard
Safetensors
trl
sft
unsloth
Generated from Trainer
Model card Files Metrics Community
xammi commited on
Commit
7f9cbd1
•
1 Parent(s): c192b56

End of training

Browse files
Files changed (8) hide show
  1. README.md +110 -50
  2. adapter_config.json +9 -9
  3. adapter_model.safetensors +1 -1
  4. runs/Sep24_17-55-06_sammie/events.out.tfevents.1727222136.sammie.531209.0 +3 -0
  5. runs/Sep24_18-00-27_sammie/events.out.tfevents.1727222475.sammie.1790.0 +3 -0
  6. runs/Sep24_18-28-37_sammie/events.out.tfevents.1727224142.sammie.1790.1 +3 -0
  7. tokenizer_config.json +1 -0
  8. training_args.bin +1 -1
README.md CHANGED
@@ -6,7 +6,7 @@ tags:
6
  - sft
7
  - unsloth
8
  - generated_from_trainer
9
- base_model: unsloth/tinyllama-bnb-4bit
10
  datasets:
11
  - generator
12
  model-index:
@@ -19,9 +19,9 @@ should probably proofread and complete it, then remove this comment. -->
19
 
20
  # MunicipalPredictionModel-TinyLlama
21
 
22
- This model is a fine-tuned version of [unsloth/tinyllama-bnb-4bit](https://huggingface.co/unsloth/tinyllama-bnb-4bit) on the generator dataset.
23
  It achieves the following results on the evaluation set:
24
- - Loss: 1.0067
25
 
26
  ## Model description
27
 
@@ -48,59 +48,119 @@ The following hyperparameters were used during training:
48
  - total_train_batch_size: 32
49
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
50
  - lr_scheduler_type: linear
51
- - num_epochs: 2
52
  - mixed_precision_training: Native AMP
53
 
54
  ### Training results
55
 
56
  | Training Loss | Epoch | Step | Validation Loss |
57
  |:-------------:|:------:|:----:|:---------------:|
58
- | 1.4976 | 0.0431 | 10 | 1.3395 |
59
- | 1.2954 | 0.0862 | 20 | 1.2420 |
60
- | 1.2219 | 0.1293 | 30 | 1.1957 |
61
- | 1.1914 | 0.1724 | 40 | 1.1660 |
62
- | 1.1515 | 0.2155 | 50 | 1.1449 |
63
- | 1.1491 | 0.2586 | 60 | 1.1301 |
64
- | 1.129 | 0.3017 | 70 | 1.1172 |
65
- | 1.1175 | 0.3448 | 80 | 1.1070 |
66
- | 1.1129 | 0.3879 | 90 | 1.0982 |
67
- | 1.088 | 0.4310 | 100 | 1.0902 |
68
- | 1.0823 | 0.4741 | 110 | 1.0847 |
69
- | 1.0912 | 0.5172 | 120 | 1.0789 |
70
- | 1.0847 | 0.5603 | 130 | 1.0734 |
71
- | 1.0773 | 0.6034 | 140 | 1.0689 |
72
- | 1.0671 | 0.6466 | 150 | 1.0632 |
73
- | 1.0711 | 0.6897 | 160 | 1.0602 |
74
- | 1.0607 | 0.7328 | 170 | 1.0558 |
75
- | 1.0579 | 0.7759 | 180 | 1.0515 |
76
- | 1.0565 | 0.8190 | 190 | 1.0481 |
77
- | 1.0605 | 0.8621 | 200 | 1.0459 |
78
- | 1.0626 | 0.9052 | 210 | 1.0413 |
79
- | 1.0363 | 0.9483 | 220 | 1.0392 |
80
- | 1.0422 | 0.9914 | 230 | 1.0363 |
81
- | 1.0365 | 1.0345 | 240 | 1.0359 |
82
- | 1.0083 | 1.0776 | 250 | 1.0327 |
83
- | 1.0138 | 1.1207 | 260 | 1.0297 |
84
- | 1.015 | 1.1638 | 270 | 1.0275 |
85
- | 1.0116 | 1.2069 | 280 | 1.0258 |
86
- | 1.0168 | 1.25 | 290 | 1.0245 |
87
- | 1.0147 | 1.2931 | 300 | 1.0225 |
88
- | 1.022 | 1.3362 | 310 | 1.0203 |
89
- | 1.0118 | 1.3793 | 320 | 1.0199 |
90
- | 1.0069 | 1.4224 | 330 | 1.0181 |
91
- | 1.0042 | 1.4655 | 340 | 1.0163 |
92
- | 0.9963 | 1.5086 | 350 | 1.0145 |
93
- | 1.0127 | 1.5517 | 360 | 1.0137 |
94
- | 1.0076 | 1.5948 | 370 | 1.0125 |
95
- | 1.0127 | 1.6379 | 380 | 1.0113 |
96
- | 1.0048 | 1.6810 | 390 | 1.0102 |
97
- | 0.9904 | 1.7241 | 400 | 1.0093 |
98
- | 0.9971 | 1.7672 | 410 | 1.0091 |
99
- | 1.0073 | 1.8103 | 420 | 1.0082 |
100
- | 1.0138 | 1.8534 | 430 | 1.0077 |
101
- | 0.9952 | 1.8966 | 440 | 1.0074 |
102
- | 0.9977 | 1.9397 | 450 | 1.0070 |
103
- | 0.9966 | 1.9828 | 460 | 1.0067 |
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
104
 
105
 
106
  ### Framework versions
 
6
  - sft
7
  - unsloth
8
  - generated_from_trainer
9
+ base_model: unsloth/tinyllama-chat-bnb-4bit
10
  datasets:
11
  - generator
12
  model-index:
 
19
 
20
  # MunicipalPredictionModel-TinyLlama
21
 
22
+ This model is a fine-tuned version of [unsloth/tinyllama-chat-bnb-4bit](https://huggingface.co/unsloth/tinyllama-chat-bnb-4bit) on the generator dataset.
23
  It achieves the following results on the evaluation set:
24
+ - Loss: 1.3256
25
 
26
  ## Model description
27
 
 
48
  - total_train_batch_size: 32
49
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
50
  - lr_scheduler_type: linear
51
+ - num_epochs: 1
52
  - mixed_precision_training: Native AMP
53
 
54
  ### Training results
55
 
56
  | Training Loss | Epoch | Step | Validation Loss |
57
  |:-------------:|:------:|:----:|:---------------:|
58
+ | 2.0636 | 0.0094 | 10 | 1.9387 |
59
+ | 1.8627 | 0.0188 | 20 | 1.8200 |
60
+ | 1.7274 | 0.0282 | 30 | 1.7460 |
61
+ | 1.7278 | 0.0377 | 40 | 1.6963 |
62
+ | 1.6922 | 0.0471 | 50 | 1.6616 |
63
+ | 1.6037 | 0.0565 | 60 | 1.6337 |
64
+ | 1.5869 | 0.0659 | 70 | 1.6158 |
65
+ | 1.5963 | 0.0753 | 80 | 1.5981 |
66
+ | 1.5748 | 0.0847 | 90 | 1.5865 |
67
+ | 1.5177 | 0.0942 | 100 | 1.5741 |
68
+ | 1.5486 | 0.1036 | 110 | 1.5612 |
69
+ | 1.5296 | 0.1130 | 120 | 1.5508 |
70
+ | 1.5375 | 0.1224 | 130 | 1.5413 |
71
+ | 1.5544 | 0.1318 | 140 | 1.5346 |
72
+ | 1.4778 | 0.1412 | 150 | 1.5259 |
73
+ | 1.5084 | 0.1507 | 160 | 1.5194 |
74
+ | 1.4812 | 0.1601 | 170 | 1.5132 |
75
+ | 1.4668 | 0.1695 | 180 | 1.5077 |
76
+ | 1.4839 | 0.1789 | 190 | 1.5023 |
77
+ | 1.4593 | 0.1883 | 200 | 1.4966 |
78
+ | 1.5284 | 0.1977 | 210 | 1.4897 |
79
+ | 1.4317 | 0.2072 | 220 | 1.4848 |
80
+ | 1.4335 | 0.2166 | 230 | 1.4771 |
81
+ | 1.4312 | 0.2260 | 240 | 1.4730 |
82
+ | 1.4437 | 0.2354 | 250 | 1.4675 |
83
+ | 1.4266 | 0.2448 | 260 | 1.4659 |
84
+ | 1.4569 | 0.2542 | 270 | 1.4604 |
85
+ | 1.4287 | 0.2637 | 280 | 1.4549 |
86
+ | 1.412 | 0.2731 | 290 | 1.4505 |
87
+ | 1.4113 | 0.2825 | 300 | 1.4474 |
88
+ | 1.4106 | 0.2919 | 310 | 1.4435 |
89
+ | 1.4074 | 0.3013 | 320 | 1.4398 |
90
+ | 1.4139 | 0.3107 | 330 | 1.4362 |
91
+ | 1.3989 | 0.3202 | 340 | 1.4325 |
92
+ | 1.3803 | 0.3296 | 350 | 1.4292 |
93
+ | 1.3489 | 0.3390 | 360 | 1.4272 |
94
+ | 1.4065 | 0.3484 | 370 | 1.4236 |
95
+ | 1.4201 | 0.3578 | 380 | 1.4208 |
96
+ | 1.3936 | 0.3672 | 390 | 1.4174 |
97
+ | 1.3934 | 0.3766 | 400 | 1.4142 |
98
+ | 1.4117 | 0.3861 | 410 | 1.4121 |
99
+ | 1.3812 | 0.3955 | 420 | 1.4091 |
100
+ | 1.3882 | 0.4049 | 430 | 1.4065 |
101
+ | 1.3849 | 0.4143 | 440 | 1.4039 |
102
+ | 1.3537 | 0.4237 | 450 | 1.4011 |
103
+ | 1.3892 | 0.4331 | 460 | 1.3993 |
104
+ | 1.3915 | 0.4426 | 470 | 1.3971 |
105
+ | 1.3766 | 0.4520 | 480 | 1.3967 |
106
+ | 1.3422 | 0.4614 | 490 | 1.3968 |
107
+ | 1.3525 | 0.4708 | 500 | 1.3917 |
108
+ | 1.3662 | 0.4802 | 510 | 1.3878 |
109
+ | 1.3254 | 0.4896 | 520 | 1.3871 |
110
+ | 1.3047 | 0.4991 | 530 | 1.3840 |
111
+ | 1.3064 | 0.5085 | 540 | 1.3820 |
112
+ | 1.3662 | 0.5179 | 550 | 1.3805 |
113
+ | 1.3177 | 0.5273 | 560 | 1.3793 |
114
+ | 1.315 | 0.5367 | 570 | 1.3762 |
115
+ | 1.3275 | 0.5461 | 580 | 1.3742 |
116
+ | 1.3631 | 0.5556 | 590 | 1.3727 |
117
+ | 1.3324 | 0.5650 | 600 | 1.3697 |
118
+ | 1.317 | 0.5744 | 610 | 1.3679 |
119
+ | 1.3479 | 0.5838 | 620 | 1.3669 |
120
+ | 1.3526 | 0.5932 | 630 | 1.3650 |
121
+ | 1.2873 | 0.6026 | 640 | 1.3640 |
122
+ | 1.3318 | 0.6121 | 650 | 1.3615 |
123
+ | 1.302 | 0.6215 | 660 | 1.3599 |
124
+ | 1.2907 | 0.6309 | 670 | 1.3582 |
125
+ | 1.3387 | 0.6403 | 680 | 1.3565 |
126
+ | 1.3278 | 0.6497 | 690 | 1.3552 |
127
+ | 1.3073 | 0.6591 | 700 | 1.3538 |
128
+ | 1.2986 | 0.6685 | 710 | 1.3532 |
129
+ | 1.3461 | 0.6780 | 720 | 1.3509 |
130
+ | 1.3229 | 0.6874 | 730 | 1.3500 |
131
+ | 1.3055 | 0.6968 | 740 | 1.3490 |
132
+ | 1.3307 | 0.7062 | 750 | 1.3476 |
133
+ | 1.3308 | 0.7156 | 760 | 1.3468 |
134
+ | 1.3083 | 0.7250 | 770 | 1.3459 |
135
+ | 1.312 | 0.7345 | 780 | 1.3447 |
136
+ | 1.3068 | 0.7439 | 790 | 1.3424 |
137
+ | 1.3202 | 0.7533 | 800 | 1.3415 |
138
+ | 1.2872 | 0.7627 | 810 | 1.3403 |
139
+ | 1.311 | 0.7721 | 820 | 1.3393 |
140
+ | 1.3213 | 0.7815 | 830 | 1.3385 |
141
+ | 1.3302 | 0.7910 | 840 | 1.3372 |
142
+ | 1.3243 | 0.8004 | 850 | 1.3362 |
143
+ | 1.3077 | 0.8098 | 860 | 1.3352 |
144
+ | 1.3229 | 0.8192 | 870 | 1.3344 |
145
+ | 1.3243 | 0.8286 | 880 | 1.3333 |
146
+ | 1.2909 | 0.8380 | 890 | 1.3327 |
147
+ | 1.3205 | 0.8475 | 900 | 1.3319 |
148
+ | 1.2863 | 0.8569 | 910 | 1.3314 |
149
+ | 1.2583 | 0.8663 | 920 | 1.3305 |
150
+ | 1.29 | 0.8757 | 930 | 1.3298 |
151
+ | 1.2706 | 0.8851 | 940 | 1.3294 |
152
+ | 1.272 | 0.8945 | 950 | 1.3291 |
153
+ | 1.2899 | 0.9040 | 960 | 1.3285 |
154
+ | 1.3184 | 0.9134 | 970 | 1.3280 |
155
+ | 1.3323 | 0.9228 | 980 | 1.3276 |
156
+ | 1.3314 | 0.9322 | 990 | 1.3272 |
157
+ | 1.2965 | 0.9416 | 1000 | 1.3269 |
158
+ | 1.2874 | 0.9510 | 1010 | 1.3266 |
159
+ | 1.2569 | 0.9605 | 1020 | 1.3263 |
160
+ | 1.2622 | 0.9699 | 1030 | 1.3261 |
161
+ | 1.3258 | 0.9793 | 1040 | 1.3259 |
162
+ | 1.3072 | 0.9887 | 1050 | 1.3257 |
163
+ | 1.2555 | 0.9981 | 1060 | 1.3256 |
164
 
165
 
166
  ### Framework versions
adapter_config.json CHANGED
@@ -1,7 +1,7 @@
1
  {
2
  "alpha_pattern": {},
3
  "auto_mapping": null,
4
- "base_model_name_or_path": "unsloth/tinyllama-bnb-4bit",
5
  "bias": "none",
6
  "fan_in_fan_out": false,
7
  "inference_mode": true,
@@ -20,17 +20,17 @@
20
  "rank_pattern": {},
21
  "revision": null,
22
  "target_modules": [
 
 
 
23
  "gate_proj",
24
- "down_proj",
25
- "mlp.gate_up_proj",
26
  "v_proj",
27
- "q_proj",
28
- "up_proj",
29
  "mlp.down_proj",
30
- "o_proj",
31
- "self_attn.qkv_proj.weight",
32
- "self_attn.o_proj.weight",
33
- "k_proj"
 
34
  ],
35
  "task_type": "CAUSAL_LM",
36
  "use_dora": false,
 
1
  {
2
  "alpha_pattern": {},
3
  "auto_mapping": null,
4
+ "base_model_name_or_path": "unsloth/tinyllama-chat-bnb-4bit",
5
  "bias": "none",
6
  "fan_in_fan_out": false,
7
  "inference_mode": true,
 
20
  "rank_pattern": {},
21
  "revision": null,
22
  "target_modules": [
23
+ "self_attn.o_proj.weight",
24
+ "up_proj",
25
+ "self_attn.qkv_proj.weight",
26
  "gate_proj",
 
 
27
  "v_proj",
 
 
28
  "mlp.down_proj",
29
+ "k_proj",
30
+ "q_proj",
31
+ "mlp.gate_up_proj",
32
+ "down_proj",
33
+ "o_proj"
34
  ],
35
  "task_type": "CAUSAL_LM",
36
  "use_dora": false,
adapter_model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:65aac8e52d2b349519b5438dcb882cfad4794a93ffd157f9b49d878172d92e61
3
  size 403743472
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:557552902b20a33a04721e891b1592661f4067c47e2d1ba906292388ef38302b
3
  size 403743472
runs/Sep24_17-55-06_sammie/events.out.tfevents.1727222136.sammie.531209.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b8a840a24889feeb3620ed25f9955c3cabdf341c51a6609fe13f30fafa174a29
3
+ size 4184
runs/Sep24_18-00-27_sammie/events.out.tfevents.1727222475.sammie.1790.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3c96541528a41c661583e6bc1022a7958a3eb297a426e4be6fe391f6c3d455f0
3
+ size 7379
runs/Sep24_18-28-37_sammie/events.out.tfevents.1727224142.sammie.1790.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cc27f179a249d0691478cd0f935f67b81c45e4c6cd42c17d0a8e1036df63caa3
3
+ size 57298
tokenizer_config.json CHANGED
@@ -29,6 +29,7 @@
29
  }
30
  },
31
  "bos_token": "<s>",
 
32
  "clean_up_tokenization_spaces": false,
33
  "eos_token": "</s>",
34
  "legacy": false,
 
29
  }
30
  },
31
  "bos_token": "<s>",
32
+ "chat_template": "{% for message in messages %}\n{% if message['role'] == 'user' %}\n{{ '<|user|>\n' + message['content'] + eos_token }}\n{% elif message['role'] == 'system' %}\n{{ '<|system|>\n' + message['content'] + eos_token }}\n{% elif message['role'] == 'assistant' %}\n{{ '<|assistant|>\n' + message['content'] + eos_token }}\n{% endif %}\n{% if loop.last and add_generation_prompt %}\n{{ '<|assistant|>' }}\n{% endif %}\n{% endfor %}",
33
  "clean_up_tokenization_spaces": false,
34
  "eos_token": "</s>",
35
  "legacy": false,
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:b191efad8e0c67e59627b59005b61e1c57ea0e8f2b81b8642ae1ab1df6697c9a
3
  size 5560
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8604a1a40ba8f678deb2830727de0c6cd6660e4c0a7d61f8a3c6b30c00e31773
3
  size 5560