Initial commit: MNLP M3 quantized model
Browse files- checkpoint-11502/config.json +73 -0
- checkpoint-11502/generation_config.json +6 -0
- checkpoint-11502/model.safetensors +3 -0
- checkpoint-11502/optimizer.pt +2 -2
- checkpoint-11502/rng_state.pth +1 -1
- checkpoint-11502/scheduler.pt +1 -1
- checkpoint-11502/trainer_state.json +78 -78
- checkpoint-11502/training_args.bin +1 -1
- checkpoint-17253/config.json +73 -0
- checkpoint-17253/generation_config.json +6 -0
- checkpoint-17253/model.safetensors +3 -0
- checkpoint-17253/optimizer.pt +2 -2
- checkpoint-17253/rng_state.pth +1 -1
- checkpoint-17253/scheduler.pt +1 -1
- checkpoint-17253/trainer_state.json +115 -115
- checkpoint-17253/training_args.bin +1 -1
- checkpoint-5751/config.json +73 -0
- checkpoint-5751/generation_config.json +6 -0
- checkpoint-5751/model.safetensors +3 -0
- checkpoint-5751/optimizer.pt +2 -2
- checkpoint-5751/rng_state.pth +1 -1
- checkpoint-5751/scheduler.pt +1 -1
- checkpoint-5751/trainer_state.json +38 -38
- checkpoint-5751/training_args.bin +1 -1
checkpoint-11502/config.json
ADDED
|
@@ -0,0 +1,73 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"Qwen3ForCausalLM"
|
| 4 |
+
],
|
| 5 |
+
"attention_bias": false,
|
| 6 |
+
"attention_dropout": 0.0,
|
| 7 |
+
"bos_token_id": 151643,
|
| 8 |
+
"eos_token_id": 151643,
|
| 9 |
+
"head_dim": 128,
|
| 10 |
+
"hidden_act": "silu",
|
| 11 |
+
"hidden_size": 1024,
|
| 12 |
+
"initializer_range": 0.02,
|
| 13 |
+
"intermediate_size": 3072,
|
| 14 |
+
"max_position_embeddings": 32768,
|
| 15 |
+
"max_window_layers": 28,
|
| 16 |
+
"model_type": "qwen3",
|
| 17 |
+
"num_attention_heads": 16,
|
| 18 |
+
"num_hidden_layers": 28,
|
| 19 |
+
"num_key_value_heads": 8,
|
| 20 |
+
"quantization_config": {
|
| 21 |
+
"config_groups": {
|
| 22 |
+
"group_0": {
|
| 23 |
+
"input_activations": {
|
| 24 |
+
"actorder": null,
|
| 25 |
+
"block_structure": null,
|
| 26 |
+
"dynamic": true,
|
| 27 |
+
"group_size": null,
|
| 28 |
+
"num_bits": 8,
|
| 29 |
+
"observer": null,
|
| 30 |
+
"observer_kwargs": {},
|
| 31 |
+
"strategy": "token",
|
| 32 |
+
"symmetric": true,
|
| 33 |
+
"type": "int"
|
| 34 |
+
},
|
| 35 |
+
"output_activations": null,
|
| 36 |
+
"targets": [
|
| 37 |
+
"Linear"
|
| 38 |
+
],
|
| 39 |
+
"weights": {
|
| 40 |
+
"actorder": null,
|
| 41 |
+
"block_structure": null,
|
| 42 |
+
"dynamic": false,
|
| 43 |
+
"group_size": null,
|
| 44 |
+
"num_bits": 8,
|
| 45 |
+
"observer": "minmax",
|
| 46 |
+
"observer_kwargs": {},
|
| 47 |
+
"strategy": "channel",
|
| 48 |
+
"symmetric": true,
|
| 49 |
+
"type": "int"
|
| 50 |
+
}
|
| 51 |
+
}
|
| 52 |
+
},
|
| 53 |
+
"format": "int-quantized",
|
| 54 |
+
"global_compression_ratio": null,
|
| 55 |
+
"ignore": [
|
| 56 |
+
"lm_head"
|
| 57 |
+
],
|
| 58 |
+
"kv_cache_scheme": null,
|
| 59 |
+
"quant_method": "compressed-tensors",
|
| 60 |
+
"quantization_status": "compressed",
|
| 61 |
+
"sparsity_config": {}
|
| 62 |
+
},
|
| 63 |
+
"rms_norm_eps": 1e-06,
|
| 64 |
+
"rope_scaling": null,
|
| 65 |
+
"rope_theta": 1000000,
|
| 66 |
+
"sliding_window": null,
|
| 67 |
+
"tie_word_embeddings": true,
|
| 68 |
+
"torch_dtype": "float16",
|
| 69 |
+
"transformers_version": "4.51.3",
|
| 70 |
+
"use_cache": true,
|
| 71 |
+
"use_sliding_window": false,
|
| 72 |
+
"vocab_size": 151936
|
| 73 |
+
}
|
checkpoint-11502/generation_config.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token_id": 151643,
|
| 3 |
+
"eos_token_id": 151643,
|
| 4 |
+
"max_new_tokens": 2048,
|
| 5 |
+
"transformers_version": "4.51.3"
|
| 6 |
+
}
|
checkpoint-11502/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bb3d1bd95c53800ef1cb04bf0b4dca4678ae38f01058094213011b8ab1df2db9
|
| 3 |
+
size 1192845752
|
checkpoint-11502/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6ea4ec19e10fcf879489137083c39125bcd2f58e8de40103fa46d5219e46c904
|
| 3 |
+
size 622337116
|
checkpoint-11502/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14244
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8d138cfe3a4adf21f048848ee35837c9a757a0a3616ff7adbb45b69aac247435
|
| 3 |
size 14244
|
checkpoint-11502/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1064
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b0ad21997940ab64708a0aee1c86735a1db628310427e22f01fbc6ae661643ac
|
| 3 |
size 1064
|
checkpoint-11502/trainer_state.json
CHANGED
|
@@ -11,179 +11,179 @@
|
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
"epoch": 0.0869414014953921,
|
| 14 |
-
"grad_norm":
|
| 15 |
-
"learning_rate": 0.
|
| 16 |
-
"loss":
|
| 17 |
"step": 500
|
| 18 |
},
|
| 19 |
{
|
| 20 |
"epoch": 0.1738828029907842,
|
| 21 |
-
"grad_norm":
|
| 22 |
-
"learning_rate": 0.
|
| 23 |
-
"loss": 0.
|
| 24 |
"step": 1000
|
| 25 |
},
|
| 26 |
{
|
| 27 |
"epoch": 0.2608242044861763,
|
| 28 |
-
"grad_norm":
|
| 29 |
-
"learning_rate": 0.
|
| 30 |
-
"loss": 0.
|
| 31 |
"step": 1500
|
| 32 |
},
|
| 33 |
{
|
| 34 |
"epoch": 0.3477656059815684,
|
| 35 |
-
"grad_norm":
|
| 36 |
-
"learning_rate": 0.
|
| 37 |
-
"loss": 0.
|
| 38 |
"step": 2000
|
| 39 |
},
|
| 40 |
{
|
| 41 |
"epoch": 0.4347070074769605,
|
| 42 |
-
"grad_norm":
|
| 43 |
-
"learning_rate": 0.
|
| 44 |
-
"loss": 0.
|
| 45 |
"step": 2500
|
| 46 |
},
|
| 47 |
{
|
| 48 |
"epoch": 0.5216484089723527,
|
| 49 |
-
"grad_norm":
|
| 50 |
-
"learning_rate": 0.
|
| 51 |
-
"loss": 0.
|
| 52 |
"step": 3000
|
| 53 |
},
|
| 54 |
{
|
| 55 |
"epoch": 0.6085898104677447,
|
| 56 |
-
"grad_norm":
|
| 57 |
-
"learning_rate": 0.
|
| 58 |
-
"loss": 0.
|
| 59 |
"step": 3500
|
| 60 |
},
|
| 61 |
{
|
| 62 |
"epoch": 0.6955312119631368,
|
| 63 |
-
"grad_norm":
|
| 64 |
-
"learning_rate": 0.
|
| 65 |
-
"loss": 0.
|
| 66 |
"step": 4000
|
| 67 |
},
|
| 68 |
{
|
| 69 |
"epoch": 0.7824726134585289,
|
| 70 |
-
"grad_norm":
|
| 71 |
-
"learning_rate": 0.
|
| 72 |
-
"loss": 0.
|
| 73 |
"step": 4500
|
| 74 |
},
|
| 75 |
{
|
| 76 |
"epoch": 0.869414014953921,
|
| 77 |
-
"grad_norm":
|
| 78 |
-
"learning_rate": 0.
|
| 79 |
-
"loss": 0.
|
| 80 |
"step": 5000
|
| 81 |
},
|
| 82 |
{
|
| 83 |
"epoch": 0.9563554164493132,
|
| 84 |
-
"grad_norm":
|
| 85 |
-
"learning_rate": 0.
|
| 86 |
-
"loss": 0.
|
| 87 |
"step": 5500
|
| 88 |
},
|
| 89 |
{
|
| 90 |
"epoch": 1.0,
|
| 91 |
-
"eval_loss":
|
| 92 |
-
"eval_runtime":
|
| 93 |
-
"eval_samples_per_second":
|
| 94 |
-
"eval_steps_per_second":
|
| 95 |
"step": 5751
|
| 96 |
},
|
| 97 |
{
|
| 98 |
"epoch": 1.0432968179447053,
|
| 99 |
-
"grad_norm":
|
| 100 |
-
"learning_rate": 0.
|
| 101 |
-
"loss": 0.
|
| 102 |
"step": 6000
|
| 103 |
},
|
| 104 |
{
|
| 105 |
"epoch": 1.1302382194400973,
|
| 106 |
-
"grad_norm":
|
| 107 |
-
"learning_rate": 0.
|
| 108 |
-
"loss": 0.
|
| 109 |
"step": 6500
|
| 110 |
},
|
| 111 |
{
|
| 112 |
"epoch": 1.2171796209354895,
|
| 113 |
-
"grad_norm":
|
| 114 |
-
"learning_rate": 0.
|
| 115 |
-
"loss": 0.
|
| 116 |
"step": 7000
|
| 117 |
},
|
| 118 |
{
|
| 119 |
"epoch": 1.3041210224308815,
|
| 120 |
-
"grad_norm":
|
| 121 |
-
"learning_rate": 0.
|
| 122 |
-
"loss": 0.
|
| 123 |
"step": 7500
|
| 124 |
},
|
| 125 |
{
|
| 126 |
"epoch": 1.3910624239262737,
|
| 127 |
-
"grad_norm":
|
| 128 |
-
"learning_rate": 0.
|
| 129 |
-
"loss": 0.
|
| 130 |
"step": 8000
|
| 131 |
},
|
| 132 |
{
|
| 133 |
"epoch": 1.4780038254216659,
|
| 134 |
-
"grad_norm":
|
| 135 |
-
"learning_rate": 0.
|
| 136 |
-
"loss": 0.
|
| 137 |
"step": 8500
|
| 138 |
},
|
| 139 |
{
|
| 140 |
"epoch": 1.5649452269170578,
|
| 141 |
-
"grad_norm":
|
| 142 |
-
"learning_rate": 9.
|
| 143 |
-
"loss": 0.
|
| 144 |
"step": 9000
|
| 145 |
},
|
| 146 |
{
|
| 147 |
"epoch": 1.65188662841245,
|
| 148 |
-
"grad_norm":
|
| 149 |
-
"learning_rate": 8.
|
| 150 |
-
"loss": 0.
|
| 151 |
"step": 9500
|
| 152 |
},
|
| 153 |
{
|
| 154 |
"epoch": 1.7388280299078422,
|
| 155 |
-
"grad_norm":
|
| 156 |
-
"learning_rate": 8.
|
| 157 |
-
"loss": 0.
|
| 158 |
"step": 10000
|
| 159 |
},
|
| 160 |
{
|
| 161 |
"epoch": 1.8257694314032342,
|
| 162 |
-
"grad_norm":
|
| 163 |
-
"learning_rate": 7.
|
| 164 |
-
"loss": 0.
|
| 165 |
"step": 10500
|
| 166 |
},
|
| 167 |
{
|
| 168 |
"epoch": 1.9127108328986262,
|
| 169 |
-
"grad_norm":
|
| 170 |
-
"learning_rate": 7.
|
| 171 |
-
"loss": 0.
|
| 172 |
"step": 11000
|
| 173 |
},
|
| 174 |
{
|
| 175 |
"epoch": 1.9996522343940184,
|
| 176 |
-
"grad_norm":
|
| 177 |
-
"learning_rate": 6.
|
| 178 |
-
"loss": 0.
|
| 179 |
"step": 11500
|
| 180 |
},
|
| 181 |
{
|
| 182 |
"epoch": 2.0,
|
| 183 |
-
"eval_loss":
|
| 184 |
-
"eval_runtime":
|
| 185 |
-
"eval_samples_per_second":
|
| 186 |
-
"eval_steps_per_second":
|
| 187 |
"step": 11502
|
| 188 |
}
|
| 189 |
],
|
|
@@ -204,7 +204,7 @@
|
|
| 204 |
"attributes": {}
|
| 205 |
}
|
| 206 |
},
|
| 207 |
-
"total_flos": 6.
|
| 208 |
"train_batch_size": 4,
|
| 209 |
"trial_name": null,
|
| 210 |
"trial_params": null
|
|
|
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
"epoch": 0.0869414014953921,
|
| 14 |
+
"grad_norm": NaN,
|
| 15 |
+
"learning_rate": 0.00019421549875383993,
|
| 16 |
+
"loss": 4.2823,
|
| 17 |
"step": 500
|
| 18 |
},
|
| 19 |
{
|
| 20 |
"epoch": 0.1738828029907842,
|
| 21 |
+
"grad_norm": NaN,
|
| 22 |
+
"learning_rate": 0.0001884194053208138,
|
| 23 |
+
"loss": 0.0,
|
| 24 |
"step": 1000
|
| 25 |
},
|
| 26 |
{
|
| 27 |
"epoch": 0.2608242044861763,
|
| 28 |
+
"grad_norm": NaN,
|
| 29 |
+
"learning_rate": 0.00018262331188778763,
|
| 30 |
+
"loss": 0.0,
|
| 31 |
"step": 1500
|
| 32 |
},
|
| 33 |
{
|
| 34 |
"epoch": 0.3477656059815684,
|
| 35 |
+
"grad_norm": NaN,
|
| 36 |
+
"learning_rate": 0.0001768272184547615,
|
| 37 |
+
"loss": 0.0,
|
| 38 |
"step": 2000
|
| 39 |
},
|
| 40 |
{
|
| 41 |
"epoch": 0.4347070074769605,
|
| 42 |
+
"grad_norm": NaN,
|
| 43 |
+
"learning_rate": 0.00017103112502173535,
|
| 44 |
+
"loss": 0.0,
|
| 45 |
"step": 2500
|
| 46 |
},
|
| 47 |
{
|
| 48 |
"epoch": 0.5216484089723527,
|
| 49 |
+
"grad_norm": NaN,
|
| 50 |
+
"learning_rate": 0.0001652350315887092,
|
| 51 |
+
"loss": 0.0,
|
| 52 |
"step": 3000
|
| 53 |
},
|
| 54 |
{
|
| 55 |
"epoch": 0.6085898104677447,
|
| 56 |
+
"grad_norm": NaN,
|
| 57 |
+
"learning_rate": 0.00015943893815568308,
|
| 58 |
+
"loss": 0.0,
|
| 59 |
"step": 3500
|
| 60 |
},
|
| 61 |
{
|
| 62 |
"epoch": 0.6955312119631368,
|
| 63 |
+
"grad_norm": NaN,
|
| 64 |
+
"learning_rate": 0.00015364284472265693,
|
| 65 |
+
"loss": 0.0,
|
| 66 |
"step": 4000
|
| 67 |
},
|
| 68 |
{
|
| 69 |
"epoch": 0.7824726134585289,
|
| 70 |
+
"grad_norm": NaN,
|
| 71 |
+
"learning_rate": 0.0001478467512896308,
|
| 72 |
+
"loss": 0.0,
|
| 73 |
"step": 4500
|
| 74 |
},
|
| 75 |
{
|
| 76 |
"epoch": 0.869414014953921,
|
| 77 |
+
"grad_norm": NaN,
|
| 78 |
+
"learning_rate": 0.00014205065785660465,
|
| 79 |
+
"loss": 0.0,
|
| 80 |
"step": 5000
|
| 81 |
},
|
| 82 |
{
|
| 83 |
"epoch": 0.9563554164493132,
|
| 84 |
+
"grad_norm": NaN,
|
| 85 |
+
"learning_rate": 0.0001362545644235785,
|
| 86 |
+
"loss": 0.0,
|
| 87 |
"step": 5500
|
| 88 |
},
|
| 89 |
{
|
| 90 |
"epoch": 1.0,
|
| 91 |
+
"eval_loss": NaN,
|
| 92 |
+
"eval_runtime": 152.5081,
|
| 93 |
+
"eval_samples_per_second": 16.76,
|
| 94 |
+
"eval_steps_per_second": 4.19,
|
| 95 |
"step": 5751
|
| 96 |
},
|
| 97 |
{
|
| 98 |
"epoch": 1.0432968179447053,
|
| 99 |
+
"grad_norm": NaN,
|
| 100 |
+
"learning_rate": 0.00013045847099055238,
|
| 101 |
+
"loss": 0.0,
|
| 102 |
"step": 6000
|
| 103 |
},
|
| 104 |
{
|
| 105 |
"epoch": 1.1302382194400973,
|
| 106 |
+
"grad_norm": NaN,
|
| 107 |
+
"learning_rate": 0.00012466237755752623,
|
| 108 |
+
"loss": 0.0,
|
| 109 |
"step": 6500
|
| 110 |
},
|
| 111 |
{
|
| 112 |
"epoch": 1.2171796209354895,
|
| 113 |
+
"grad_norm": NaN,
|
| 114 |
+
"learning_rate": 0.00011886628412450009,
|
| 115 |
+
"loss": 0.0,
|
| 116 |
"step": 7000
|
| 117 |
},
|
| 118 |
{
|
| 119 |
"epoch": 1.3041210224308815,
|
| 120 |
+
"grad_norm": NaN,
|
| 121 |
+
"learning_rate": 0.00011307019069147395,
|
| 122 |
+
"loss": 0.0,
|
| 123 |
"step": 7500
|
| 124 |
},
|
| 125 |
{
|
| 126 |
"epoch": 1.3910624239262737,
|
| 127 |
+
"grad_norm": NaN,
|
| 128 |
+
"learning_rate": 0.00010727409725844781,
|
| 129 |
+
"loss": 0.0,
|
| 130 |
"step": 8000
|
| 131 |
},
|
| 132 |
{
|
| 133 |
"epoch": 1.4780038254216659,
|
| 134 |
+
"grad_norm": NaN,
|
| 135 |
+
"learning_rate": 0.00010147800382542168,
|
| 136 |
+
"loss": 0.0,
|
| 137 |
"step": 8500
|
| 138 |
},
|
| 139 |
{
|
| 140 |
"epoch": 1.5649452269170578,
|
| 141 |
+
"grad_norm": NaN,
|
| 142 |
+
"learning_rate": 9.568191039239554e-05,
|
| 143 |
+
"loss": 0.0,
|
| 144 |
"step": 9000
|
| 145 |
},
|
| 146 |
{
|
| 147 |
"epoch": 1.65188662841245,
|
| 148 |
+
"grad_norm": NaN,
|
| 149 |
+
"learning_rate": 8.988581695936939e-05,
|
| 150 |
+
"loss": 0.0,
|
| 151 |
"step": 9500
|
| 152 |
},
|
| 153 |
{
|
| 154 |
"epoch": 1.7388280299078422,
|
| 155 |
+
"grad_norm": NaN,
|
| 156 |
+
"learning_rate": 8.408972352634325e-05,
|
| 157 |
+
"loss": 0.0,
|
| 158 |
"step": 10000
|
| 159 |
},
|
| 160 |
{
|
| 161 |
"epoch": 1.8257694314032342,
|
| 162 |
+
"grad_norm": NaN,
|
| 163 |
+
"learning_rate": 7.82936300933171e-05,
|
| 164 |
+
"loss": 0.0,
|
| 165 |
"step": 10500
|
| 166 |
},
|
| 167 |
{
|
| 168 |
"epoch": 1.9127108328986262,
|
| 169 |
+
"grad_norm": NaN,
|
| 170 |
+
"learning_rate": 7.249753666029096e-05,
|
| 171 |
+
"loss": 0.0,
|
| 172 |
"step": 11000
|
| 173 |
},
|
| 174 |
{
|
| 175 |
"epoch": 1.9996522343940184,
|
| 176 |
+
"grad_norm": NaN,
|
| 177 |
+
"learning_rate": 6.670144322726482e-05,
|
| 178 |
+
"loss": 0.0,
|
| 179 |
"step": 11500
|
| 180 |
},
|
| 181 |
{
|
| 182 |
"epoch": 2.0,
|
| 183 |
+
"eval_loss": NaN,
|
| 184 |
+
"eval_runtime": 152.3781,
|
| 185 |
+
"eval_samples_per_second": 16.774,
|
| 186 |
+
"eval_steps_per_second": 4.194,
|
| 187 |
"step": 11502
|
| 188 |
}
|
| 189 |
],
|
|
|
|
| 204 |
"attributes": {}
|
| 205 |
}
|
| 206 |
},
|
| 207 |
+
"total_flos": 6.230008255217664e+16,
|
| 208 |
"train_batch_size": 4,
|
| 209 |
"trial_name": null,
|
| 210 |
"trial_params": null
|
checkpoint-11502/training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5240
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:99fccf4927862cc4fd8283f3958f1ef8528c86594c8ff39225b274adde4416df
|
| 3 |
size 5240
|
checkpoint-17253/config.json
ADDED
|
@@ -0,0 +1,73 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"Qwen3ForCausalLM"
|
| 4 |
+
],
|
| 5 |
+
"attention_bias": false,
|
| 6 |
+
"attention_dropout": 0.0,
|
| 7 |
+
"bos_token_id": 151643,
|
| 8 |
+
"eos_token_id": 151643,
|
| 9 |
+
"head_dim": 128,
|
| 10 |
+
"hidden_act": "silu",
|
| 11 |
+
"hidden_size": 1024,
|
| 12 |
+
"initializer_range": 0.02,
|
| 13 |
+
"intermediate_size": 3072,
|
| 14 |
+
"max_position_embeddings": 32768,
|
| 15 |
+
"max_window_layers": 28,
|
| 16 |
+
"model_type": "qwen3",
|
| 17 |
+
"num_attention_heads": 16,
|
| 18 |
+
"num_hidden_layers": 28,
|
| 19 |
+
"num_key_value_heads": 8,
|
| 20 |
+
"quantization_config": {
|
| 21 |
+
"config_groups": {
|
| 22 |
+
"group_0": {
|
| 23 |
+
"input_activations": {
|
| 24 |
+
"actorder": null,
|
| 25 |
+
"block_structure": null,
|
| 26 |
+
"dynamic": true,
|
| 27 |
+
"group_size": null,
|
| 28 |
+
"num_bits": 8,
|
| 29 |
+
"observer": null,
|
| 30 |
+
"observer_kwargs": {},
|
| 31 |
+
"strategy": "token",
|
| 32 |
+
"symmetric": true,
|
| 33 |
+
"type": "int"
|
| 34 |
+
},
|
| 35 |
+
"output_activations": null,
|
| 36 |
+
"targets": [
|
| 37 |
+
"Linear"
|
| 38 |
+
],
|
| 39 |
+
"weights": {
|
| 40 |
+
"actorder": null,
|
| 41 |
+
"block_structure": null,
|
| 42 |
+
"dynamic": false,
|
| 43 |
+
"group_size": null,
|
| 44 |
+
"num_bits": 8,
|
| 45 |
+
"observer": "minmax",
|
| 46 |
+
"observer_kwargs": {},
|
| 47 |
+
"strategy": "channel",
|
| 48 |
+
"symmetric": true,
|
| 49 |
+
"type": "int"
|
| 50 |
+
}
|
| 51 |
+
}
|
| 52 |
+
},
|
| 53 |
+
"format": "int-quantized",
|
| 54 |
+
"global_compression_ratio": null,
|
| 55 |
+
"ignore": [
|
| 56 |
+
"lm_head"
|
| 57 |
+
],
|
| 58 |
+
"kv_cache_scheme": null,
|
| 59 |
+
"quant_method": "compressed-tensors",
|
| 60 |
+
"quantization_status": "compressed",
|
| 61 |
+
"sparsity_config": {}
|
| 62 |
+
},
|
| 63 |
+
"rms_norm_eps": 1e-06,
|
| 64 |
+
"rope_scaling": null,
|
| 65 |
+
"rope_theta": 1000000,
|
| 66 |
+
"sliding_window": null,
|
| 67 |
+
"tie_word_embeddings": true,
|
| 68 |
+
"torch_dtype": "float16",
|
| 69 |
+
"transformers_version": "4.51.3",
|
| 70 |
+
"use_cache": true,
|
| 71 |
+
"use_sliding_window": false,
|
| 72 |
+
"vocab_size": 151936
|
| 73 |
+
}
|
checkpoint-17253/generation_config.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token_id": 151643,
|
| 3 |
+
"eos_token_id": 151643,
|
| 4 |
+
"max_new_tokens": 2048,
|
| 5 |
+
"transformers_version": "4.51.3"
|
| 6 |
+
}
|
checkpoint-17253/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bb3d1bd95c53800ef1cb04bf0b4dca4678ae38f01058094213011b8ab1df2db9
|
| 3 |
+
size 1192845752
|
checkpoint-17253/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2284b8fb6f020a6f67ebde200adfab3395d52ae16e59e099cedfaf0891362079
|
| 3 |
+
size 622337116
|
checkpoint-17253/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14244
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c062f7f375beded48b5337f5a3f3a5cb38807fa3e85dbf3e294c0ab6b627bfc2
|
| 3 |
size 14244
|
checkpoint-17253/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1064
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:69d2cf8d68dbd5ab076c36e56ef45fa3ee927d877ac4ed725d8b92e0458ccf6c
|
| 3 |
size 1064
|
checkpoint-17253/trainer_state.json
CHANGED
|
@@ -11,264 +11,264 @@
|
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
"epoch": 0.0869414014953921,
|
| 14 |
-
"grad_norm":
|
| 15 |
-
"learning_rate": 0.
|
| 16 |
-
"loss":
|
| 17 |
"step": 500
|
| 18 |
},
|
| 19 |
{
|
| 20 |
"epoch": 0.1738828029907842,
|
| 21 |
-
"grad_norm":
|
| 22 |
-
"learning_rate": 0.
|
| 23 |
-
"loss": 0.
|
| 24 |
"step": 1000
|
| 25 |
},
|
| 26 |
{
|
| 27 |
"epoch": 0.2608242044861763,
|
| 28 |
-
"grad_norm":
|
| 29 |
-
"learning_rate": 0.
|
| 30 |
-
"loss": 0.
|
| 31 |
"step": 1500
|
| 32 |
},
|
| 33 |
{
|
| 34 |
"epoch": 0.3477656059815684,
|
| 35 |
-
"grad_norm":
|
| 36 |
-
"learning_rate": 0.
|
| 37 |
-
"loss": 0.
|
| 38 |
"step": 2000
|
| 39 |
},
|
| 40 |
{
|
| 41 |
"epoch": 0.4347070074769605,
|
| 42 |
-
"grad_norm":
|
| 43 |
-
"learning_rate": 0.
|
| 44 |
-
"loss": 0.
|
| 45 |
"step": 2500
|
| 46 |
},
|
| 47 |
{
|
| 48 |
"epoch": 0.5216484089723527,
|
| 49 |
-
"grad_norm":
|
| 50 |
-
"learning_rate": 0.
|
| 51 |
-
"loss": 0.
|
| 52 |
"step": 3000
|
| 53 |
},
|
| 54 |
{
|
| 55 |
"epoch": 0.6085898104677447,
|
| 56 |
-
"grad_norm":
|
| 57 |
-
"learning_rate": 0.
|
| 58 |
-
"loss": 0.
|
| 59 |
"step": 3500
|
| 60 |
},
|
| 61 |
{
|
| 62 |
"epoch": 0.6955312119631368,
|
| 63 |
-
"grad_norm":
|
| 64 |
-
"learning_rate": 0.
|
| 65 |
-
"loss": 0.
|
| 66 |
"step": 4000
|
| 67 |
},
|
| 68 |
{
|
| 69 |
"epoch": 0.7824726134585289,
|
| 70 |
-
"grad_norm":
|
| 71 |
-
"learning_rate": 0.
|
| 72 |
-
"loss": 0.
|
| 73 |
"step": 4500
|
| 74 |
},
|
| 75 |
{
|
| 76 |
"epoch": 0.869414014953921,
|
| 77 |
-
"grad_norm":
|
| 78 |
-
"learning_rate": 0.
|
| 79 |
-
"loss": 0.
|
| 80 |
"step": 5000
|
| 81 |
},
|
| 82 |
{
|
| 83 |
"epoch": 0.9563554164493132,
|
| 84 |
-
"grad_norm":
|
| 85 |
-
"learning_rate": 0.
|
| 86 |
-
"loss": 0.
|
| 87 |
"step": 5500
|
| 88 |
},
|
| 89 |
{
|
| 90 |
"epoch": 1.0,
|
| 91 |
-
"eval_loss":
|
| 92 |
-
"eval_runtime":
|
| 93 |
-
"eval_samples_per_second":
|
| 94 |
-
"eval_steps_per_second":
|
| 95 |
"step": 5751
|
| 96 |
},
|
| 97 |
{
|
| 98 |
"epoch": 1.0432968179447053,
|
| 99 |
-
"grad_norm":
|
| 100 |
-
"learning_rate": 0.
|
| 101 |
-
"loss": 0.
|
| 102 |
"step": 6000
|
| 103 |
},
|
| 104 |
{
|
| 105 |
"epoch": 1.1302382194400973,
|
| 106 |
-
"grad_norm":
|
| 107 |
-
"learning_rate": 0.
|
| 108 |
-
"loss": 0.
|
| 109 |
"step": 6500
|
| 110 |
},
|
| 111 |
{
|
| 112 |
"epoch": 1.2171796209354895,
|
| 113 |
-
"grad_norm":
|
| 114 |
-
"learning_rate": 0.
|
| 115 |
-
"loss": 0.
|
| 116 |
"step": 7000
|
| 117 |
},
|
| 118 |
{
|
| 119 |
"epoch": 1.3041210224308815,
|
| 120 |
-
"grad_norm":
|
| 121 |
-
"learning_rate": 0.
|
| 122 |
-
"loss": 0.
|
| 123 |
"step": 7500
|
| 124 |
},
|
| 125 |
{
|
| 126 |
"epoch": 1.3910624239262737,
|
| 127 |
-
"grad_norm":
|
| 128 |
-
"learning_rate": 0.
|
| 129 |
-
"loss": 0.
|
| 130 |
"step": 8000
|
| 131 |
},
|
| 132 |
{
|
| 133 |
"epoch": 1.4780038254216659,
|
| 134 |
-
"grad_norm":
|
| 135 |
-
"learning_rate": 0.
|
| 136 |
-
"loss": 0.
|
| 137 |
"step": 8500
|
| 138 |
},
|
| 139 |
{
|
| 140 |
"epoch": 1.5649452269170578,
|
| 141 |
-
"grad_norm":
|
| 142 |
-
"learning_rate": 9.
|
| 143 |
-
"loss": 0.
|
| 144 |
"step": 9000
|
| 145 |
},
|
| 146 |
{
|
| 147 |
"epoch": 1.65188662841245,
|
| 148 |
-
"grad_norm":
|
| 149 |
-
"learning_rate": 8.
|
| 150 |
-
"loss": 0.
|
| 151 |
"step": 9500
|
| 152 |
},
|
| 153 |
{
|
| 154 |
"epoch": 1.7388280299078422,
|
| 155 |
-
"grad_norm":
|
| 156 |
-
"learning_rate": 8.
|
| 157 |
-
"loss": 0.
|
| 158 |
"step": 10000
|
| 159 |
},
|
| 160 |
{
|
| 161 |
"epoch": 1.8257694314032342,
|
| 162 |
-
"grad_norm":
|
| 163 |
-
"learning_rate": 7.
|
| 164 |
-
"loss": 0.
|
| 165 |
"step": 10500
|
| 166 |
},
|
| 167 |
{
|
| 168 |
"epoch": 1.9127108328986262,
|
| 169 |
-
"grad_norm":
|
| 170 |
-
"learning_rate": 7.
|
| 171 |
-
"loss": 0.
|
| 172 |
"step": 11000
|
| 173 |
},
|
| 174 |
{
|
| 175 |
"epoch": 1.9996522343940184,
|
| 176 |
-
"grad_norm":
|
| 177 |
-
"learning_rate": 6.
|
| 178 |
-
"loss": 0.
|
| 179 |
"step": 11500
|
| 180 |
},
|
| 181 |
{
|
| 182 |
"epoch": 2.0,
|
| 183 |
-
"eval_loss":
|
| 184 |
-
"eval_runtime":
|
| 185 |
-
"eval_samples_per_second":
|
| 186 |
-
"eval_steps_per_second":
|
| 187 |
"step": 11502
|
| 188 |
},
|
| 189 |
{
|
| 190 |
"epoch": 2.0865936358894106,
|
| 191 |
-
"grad_norm":
|
| 192 |
-
"learning_rate": 6.
|
| 193 |
-
"loss": 0.
|
| 194 |
"step": 12000
|
| 195 |
},
|
| 196 |
{
|
| 197 |
"epoch": 2.1735350373848026,
|
| 198 |
-
"grad_norm":
|
| 199 |
-
"learning_rate": 5.
|
| 200 |
-
"loss": 0.
|
| 201 |
"step": 12500
|
| 202 |
},
|
| 203 |
{
|
| 204 |
"epoch": 2.2604764388801946,
|
| 205 |
-
"grad_norm":
|
| 206 |
-
"learning_rate": 4.
|
| 207 |
-
"loss": 0.
|
| 208 |
"step": 13000
|
| 209 |
},
|
| 210 |
{
|
| 211 |
"epoch": 2.347417840375587,
|
| 212 |
-
"grad_norm":
|
| 213 |
-
"learning_rate": 4.
|
| 214 |
-
"loss": 0.
|
| 215 |
"step": 13500
|
| 216 |
},
|
| 217 |
{
|
| 218 |
"epoch": 2.434359241870979,
|
| 219 |
-
"grad_norm":
|
| 220 |
-
"learning_rate": 3.
|
| 221 |
-
"loss": 0.
|
| 222 |
"step": 14000
|
| 223 |
},
|
| 224 |
{
|
| 225 |
"epoch": 2.521300643366371,
|
| 226 |
-
"grad_norm":
|
| 227 |
-
"learning_rate": 3.
|
| 228 |
-
"loss": 0.
|
| 229 |
"step": 14500
|
| 230 |
},
|
| 231 |
{
|
| 232 |
"epoch": 2.608242044861763,
|
| 233 |
-
"grad_norm":
|
| 234 |
-
"learning_rate": 2.
|
| 235 |
-
"loss": 0.
|
| 236 |
"step": 15000
|
| 237 |
},
|
| 238 |
{
|
| 239 |
"epoch": 2.6951834463571553,
|
| 240 |
-
"grad_norm":
|
| 241 |
-
"learning_rate": 2.
|
| 242 |
-
"loss": 0.
|
| 243 |
"step": 15500
|
| 244 |
},
|
| 245 |
{
|
| 246 |
"epoch": 2.7821248478525473,
|
| 247 |
-
"grad_norm":
|
| 248 |
-
"learning_rate": 1.
|
| 249 |
-
"loss": 0.
|
| 250 |
"step": 16000
|
| 251 |
},
|
| 252 |
{
|
| 253 |
"epoch": 2.8690662493479397,
|
| 254 |
-
"grad_norm":
|
| 255 |
-
"learning_rate": 8.
|
| 256 |
-
"loss": 0.
|
| 257 |
"step": 16500
|
| 258 |
},
|
| 259 |
{
|
| 260 |
"epoch": 2.9560076508433317,
|
| 261 |
-
"grad_norm":
|
| 262 |
-
"learning_rate":
|
| 263 |
-
"loss": 0.
|
| 264 |
"step": 17000
|
| 265 |
},
|
| 266 |
{
|
| 267 |
"epoch": 3.0,
|
| 268 |
-
"eval_loss":
|
| 269 |
-
"eval_runtime":
|
| 270 |
-
"eval_samples_per_second":
|
| 271 |
-
"eval_steps_per_second":
|
| 272 |
"step": 17253
|
| 273 |
}
|
| 274 |
],
|
|
@@ -289,7 +289,7 @@
|
|
| 289 |
"attributes": {}
|
| 290 |
}
|
| 291 |
},
|
| 292 |
-
"total_flos": 9.
|
| 293 |
"train_batch_size": 4,
|
| 294 |
"trial_name": null,
|
| 295 |
"trial_params": null
|
|
|
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
"epoch": 0.0869414014953921,
|
| 14 |
+
"grad_norm": NaN,
|
| 15 |
+
"learning_rate": 0.00019421549875383993,
|
| 16 |
+
"loss": 4.2823,
|
| 17 |
"step": 500
|
| 18 |
},
|
| 19 |
{
|
| 20 |
"epoch": 0.1738828029907842,
|
| 21 |
+
"grad_norm": NaN,
|
| 22 |
+
"learning_rate": 0.0001884194053208138,
|
| 23 |
+
"loss": 0.0,
|
| 24 |
"step": 1000
|
| 25 |
},
|
| 26 |
{
|
| 27 |
"epoch": 0.2608242044861763,
|
| 28 |
+
"grad_norm": NaN,
|
| 29 |
+
"learning_rate": 0.00018262331188778763,
|
| 30 |
+
"loss": 0.0,
|
| 31 |
"step": 1500
|
| 32 |
},
|
| 33 |
{
|
| 34 |
"epoch": 0.3477656059815684,
|
| 35 |
+
"grad_norm": NaN,
|
| 36 |
+
"learning_rate": 0.0001768272184547615,
|
| 37 |
+
"loss": 0.0,
|
| 38 |
"step": 2000
|
| 39 |
},
|
| 40 |
{
|
| 41 |
"epoch": 0.4347070074769605,
|
| 42 |
+
"grad_norm": NaN,
|
| 43 |
+
"learning_rate": 0.00017103112502173535,
|
| 44 |
+
"loss": 0.0,
|
| 45 |
"step": 2500
|
| 46 |
},
|
| 47 |
{
|
| 48 |
"epoch": 0.5216484089723527,
|
| 49 |
+
"grad_norm": NaN,
|
| 50 |
+
"learning_rate": 0.0001652350315887092,
|
| 51 |
+
"loss": 0.0,
|
| 52 |
"step": 3000
|
| 53 |
},
|
| 54 |
{
|
| 55 |
"epoch": 0.6085898104677447,
|
| 56 |
+
"grad_norm": NaN,
|
| 57 |
+
"learning_rate": 0.00015943893815568308,
|
| 58 |
+
"loss": 0.0,
|
| 59 |
"step": 3500
|
| 60 |
},
|
| 61 |
{
|
| 62 |
"epoch": 0.6955312119631368,
|
| 63 |
+
"grad_norm": NaN,
|
| 64 |
+
"learning_rate": 0.00015364284472265693,
|
| 65 |
+
"loss": 0.0,
|
| 66 |
"step": 4000
|
| 67 |
},
|
| 68 |
{
|
| 69 |
"epoch": 0.7824726134585289,
|
| 70 |
+
"grad_norm": NaN,
|
| 71 |
+
"learning_rate": 0.0001478467512896308,
|
| 72 |
+
"loss": 0.0,
|
| 73 |
"step": 4500
|
| 74 |
},
|
| 75 |
{
|
| 76 |
"epoch": 0.869414014953921,
|
| 77 |
+
"grad_norm": NaN,
|
| 78 |
+
"learning_rate": 0.00014205065785660465,
|
| 79 |
+
"loss": 0.0,
|
| 80 |
"step": 5000
|
| 81 |
},
|
| 82 |
{
|
| 83 |
"epoch": 0.9563554164493132,
|
| 84 |
+
"grad_norm": NaN,
|
| 85 |
+
"learning_rate": 0.0001362545644235785,
|
| 86 |
+
"loss": 0.0,
|
| 87 |
"step": 5500
|
| 88 |
},
|
| 89 |
{
|
| 90 |
"epoch": 1.0,
|
| 91 |
+
"eval_loss": NaN,
|
| 92 |
+
"eval_runtime": 152.5081,
|
| 93 |
+
"eval_samples_per_second": 16.76,
|
| 94 |
+
"eval_steps_per_second": 4.19,
|
| 95 |
"step": 5751
|
| 96 |
},
|
| 97 |
{
|
| 98 |
"epoch": 1.0432968179447053,
|
| 99 |
+
"grad_norm": NaN,
|
| 100 |
+
"learning_rate": 0.00013045847099055238,
|
| 101 |
+
"loss": 0.0,
|
| 102 |
"step": 6000
|
| 103 |
},
|
| 104 |
{
|
| 105 |
"epoch": 1.1302382194400973,
|
| 106 |
+
"grad_norm": NaN,
|
| 107 |
+
"learning_rate": 0.00012466237755752623,
|
| 108 |
+
"loss": 0.0,
|
| 109 |
"step": 6500
|
| 110 |
},
|
| 111 |
{
|
| 112 |
"epoch": 1.2171796209354895,
|
| 113 |
+
"grad_norm": NaN,
|
| 114 |
+
"learning_rate": 0.00011886628412450009,
|
| 115 |
+
"loss": 0.0,
|
| 116 |
"step": 7000
|
| 117 |
},
|
| 118 |
{
|
| 119 |
"epoch": 1.3041210224308815,
|
| 120 |
+
"grad_norm": NaN,
|
| 121 |
+
"learning_rate": 0.00011307019069147395,
|
| 122 |
+
"loss": 0.0,
|
| 123 |
"step": 7500
|
| 124 |
},
|
| 125 |
{
|
| 126 |
"epoch": 1.3910624239262737,
|
| 127 |
+
"grad_norm": NaN,
|
| 128 |
+
"learning_rate": 0.00010727409725844781,
|
| 129 |
+
"loss": 0.0,
|
| 130 |
"step": 8000
|
| 131 |
},
|
| 132 |
{
|
| 133 |
"epoch": 1.4780038254216659,
|
| 134 |
+
"grad_norm": NaN,
|
| 135 |
+
"learning_rate": 0.00010147800382542168,
|
| 136 |
+
"loss": 0.0,
|
| 137 |
"step": 8500
|
| 138 |
},
|
| 139 |
{
|
| 140 |
"epoch": 1.5649452269170578,
|
| 141 |
+
"grad_norm": NaN,
|
| 142 |
+
"learning_rate": 9.568191039239554e-05,
|
| 143 |
+
"loss": 0.0,
|
| 144 |
"step": 9000
|
| 145 |
},
|
| 146 |
{
|
| 147 |
"epoch": 1.65188662841245,
|
| 148 |
+
"grad_norm": NaN,
|
| 149 |
+
"learning_rate": 8.988581695936939e-05,
|
| 150 |
+
"loss": 0.0,
|
| 151 |
"step": 9500
|
| 152 |
},
|
| 153 |
{
|
| 154 |
"epoch": 1.7388280299078422,
|
| 155 |
+
"grad_norm": NaN,
|
| 156 |
+
"learning_rate": 8.408972352634325e-05,
|
| 157 |
+
"loss": 0.0,
|
| 158 |
"step": 10000
|
| 159 |
},
|
| 160 |
{
|
| 161 |
"epoch": 1.8257694314032342,
|
| 162 |
+
"grad_norm": NaN,
|
| 163 |
+
"learning_rate": 7.82936300933171e-05,
|
| 164 |
+
"loss": 0.0,
|
| 165 |
"step": 10500
|
| 166 |
},
|
| 167 |
{
|
| 168 |
"epoch": 1.9127108328986262,
|
| 169 |
+
"grad_norm": NaN,
|
| 170 |
+
"learning_rate": 7.249753666029096e-05,
|
| 171 |
+
"loss": 0.0,
|
| 172 |
"step": 11000
|
| 173 |
},
|
| 174 |
{
|
| 175 |
"epoch": 1.9996522343940184,
|
| 176 |
+
"grad_norm": NaN,
|
| 177 |
+
"learning_rate": 6.670144322726482e-05,
|
| 178 |
+
"loss": 0.0,
|
| 179 |
"step": 11500
|
| 180 |
},
|
| 181 |
{
|
| 182 |
"epoch": 2.0,
|
| 183 |
+
"eval_loss": NaN,
|
| 184 |
+
"eval_runtime": 152.3781,
|
| 185 |
+
"eval_samples_per_second": 16.774,
|
| 186 |
+
"eval_steps_per_second": 4.194,
|
| 187 |
"step": 11502
|
| 188 |
},
|
| 189 |
{
|
| 190 |
"epoch": 2.0865936358894106,
|
| 191 |
+
"grad_norm": NaN,
|
| 192 |
+
"learning_rate": 6.0905349794238687e-05,
|
| 193 |
+
"loss": 0.0,
|
| 194 |
"step": 12000
|
| 195 |
},
|
| 196 |
{
|
| 197 |
"epoch": 2.1735350373848026,
|
| 198 |
+
"grad_norm": NaN,
|
| 199 |
+
"learning_rate": 5.510925636121255e-05,
|
| 200 |
+
"loss": 0.0,
|
| 201 |
"step": 12500
|
| 202 |
},
|
| 203 |
{
|
| 204 |
"epoch": 2.2604764388801946,
|
| 205 |
+
"grad_norm": NaN,
|
| 206 |
+
"learning_rate": 4.9313162928186405e-05,
|
| 207 |
+
"loss": 0.0,
|
| 208 |
"step": 13000
|
| 209 |
},
|
| 210 |
{
|
| 211 |
"epoch": 2.347417840375587,
|
| 212 |
+
"grad_norm": NaN,
|
| 213 |
+
"learning_rate": 4.351706949516026e-05,
|
| 214 |
+
"loss": 0.0,
|
| 215 |
"step": 13500
|
| 216 |
},
|
| 217 |
{
|
| 218 |
"epoch": 2.434359241870979,
|
| 219 |
+
"grad_norm": NaN,
|
| 220 |
+
"learning_rate": 3.7720976062134123e-05,
|
| 221 |
+
"loss": 0.0,
|
| 222 |
"step": 14000
|
| 223 |
},
|
| 224 |
{
|
| 225 |
"epoch": 2.521300643366371,
|
| 226 |
+
"grad_norm": NaN,
|
| 227 |
+
"learning_rate": 3.1924882629107986e-05,
|
| 228 |
+
"loss": 0.0,
|
| 229 |
"step": 14500
|
| 230 |
},
|
| 231 |
{
|
| 232 |
"epoch": 2.608242044861763,
|
| 233 |
+
"grad_norm": NaN,
|
| 234 |
+
"learning_rate": 2.6128789196081842e-05,
|
| 235 |
+
"loss": 0.0,
|
| 236 |
"step": 15000
|
| 237 |
},
|
| 238 |
{
|
| 239 |
"epoch": 2.6951834463571553,
|
| 240 |
+
"grad_norm": NaN,
|
| 241 |
+
"learning_rate": 2.03326957630557e-05,
|
| 242 |
+
"loss": 0.0,
|
| 243 |
"step": 15500
|
| 244 |
},
|
| 245 |
{
|
| 246 |
"epoch": 2.7821248478525473,
|
| 247 |
+
"grad_norm": NaN,
|
| 248 |
+
"learning_rate": 1.4536602330029562e-05,
|
| 249 |
+
"loss": 0.0,
|
| 250 |
"step": 16000
|
| 251 |
},
|
| 252 |
{
|
| 253 |
"epoch": 2.8690662493479397,
|
| 254 |
+
"grad_norm": NaN,
|
| 255 |
+
"learning_rate": 8.74050889700342e-06,
|
| 256 |
+
"loss": 0.0,
|
| 257 |
"step": 16500
|
| 258 |
},
|
| 259 |
{
|
| 260 |
"epoch": 2.9560076508433317,
|
| 261 |
+
"grad_norm": NaN,
|
| 262 |
+
"learning_rate": 2.9444154639772795e-06,
|
| 263 |
+
"loss": 0.0,
|
| 264 |
"step": 17000
|
| 265 |
},
|
| 266 |
{
|
| 267 |
"epoch": 3.0,
|
| 268 |
+
"eval_loss": NaN,
|
| 269 |
+
"eval_runtime": 152.6977,
|
| 270 |
+
"eval_samples_per_second": 16.739,
|
| 271 |
+
"eval_steps_per_second": 4.185,
|
| 272 |
"step": 17253
|
| 273 |
}
|
| 274 |
],
|
|
|
|
| 289 |
"attributes": {}
|
| 290 |
}
|
| 291 |
},
|
| 292 |
+
"total_flos": 9.345012382826496e+16,
|
| 293 |
"train_batch_size": 4,
|
| 294 |
"trial_name": null,
|
| 295 |
"trial_params": null
|
checkpoint-17253/training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5240
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:99fccf4927862cc4fd8283f3958f1ef8528c86594c8ff39225b274adde4416df
|
| 3 |
size 5240
|
checkpoint-5751/config.json
ADDED
|
@@ -0,0 +1,73 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"Qwen3ForCausalLM"
|
| 4 |
+
],
|
| 5 |
+
"attention_bias": false,
|
| 6 |
+
"attention_dropout": 0.0,
|
| 7 |
+
"bos_token_id": 151643,
|
| 8 |
+
"eos_token_id": 151643,
|
| 9 |
+
"head_dim": 128,
|
| 10 |
+
"hidden_act": "silu",
|
| 11 |
+
"hidden_size": 1024,
|
| 12 |
+
"initializer_range": 0.02,
|
| 13 |
+
"intermediate_size": 3072,
|
| 14 |
+
"max_position_embeddings": 32768,
|
| 15 |
+
"max_window_layers": 28,
|
| 16 |
+
"model_type": "qwen3",
|
| 17 |
+
"num_attention_heads": 16,
|
| 18 |
+
"num_hidden_layers": 28,
|
| 19 |
+
"num_key_value_heads": 8,
|
| 20 |
+
"quantization_config": {
|
| 21 |
+
"config_groups": {
|
| 22 |
+
"group_0": {
|
| 23 |
+
"input_activations": {
|
| 24 |
+
"actorder": null,
|
| 25 |
+
"block_structure": null,
|
| 26 |
+
"dynamic": true,
|
| 27 |
+
"group_size": null,
|
| 28 |
+
"num_bits": 8,
|
| 29 |
+
"observer": null,
|
| 30 |
+
"observer_kwargs": {},
|
| 31 |
+
"strategy": "token",
|
| 32 |
+
"symmetric": true,
|
| 33 |
+
"type": "int"
|
| 34 |
+
},
|
| 35 |
+
"output_activations": null,
|
| 36 |
+
"targets": [
|
| 37 |
+
"Linear"
|
| 38 |
+
],
|
| 39 |
+
"weights": {
|
| 40 |
+
"actorder": null,
|
| 41 |
+
"block_structure": null,
|
| 42 |
+
"dynamic": false,
|
| 43 |
+
"group_size": null,
|
| 44 |
+
"num_bits": 8,
|
| 45 |
+
"observer": "minmax",
|
| 46 |
+
"observer_kwargs": {},
|
| 47 |
+
"strategy": "channel",
|
| 48 |
+
"symmetric": true,
|
| 49 |
+
"type": "int"
|
| 50 |
+
}
|
| 51 |
+
}
|
| 52 |
+
},
|
| 53 |
+
"format": "int-quantized",
|
| 54 |
+
"global_compression_ratio": null,
|
| 55 |
+
"ignore": [
|
| 56 |
+
"lm_head"
|
| 57 |
+
],
|
| 58 |
+
"kv_cache_scheme": null,
|
| 59 |
+
"quant_method": "compressed-tensors",
|
| 60 |
+
"quantization_status": "compressed",
|
| 61 |
+
"sparsity_config": {}
|
| 62 |
+
},
|
| 63 |
+
"rms_norm_eps": 1e-06,
|
| 64 |
+
"rope_scaling": null,
|
| 65 |
+
"rope_theta": 1000000,
|
| 66 |
+
"sliding_window": null,
|
| 67 |
+
"tie_word_embeddings": true,
|
| 68 |
+
"torch_dtype": "float16",
|
| 69 |
+
"transformers_version": "4.51.3",
|
| 70 |
+
"use_cache": true,
|
| 71 |
+
"use_sliding_window": false,
|
| 72 |
+
"vocab_size": 151936
|
| 73 |
+
}
|
checkpoint-5751/generation_config.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token_id": 151643,
|
| 3 |
+
"eos_token_id": 151643,
|
| 4 |
+
"max_new_tokens": 2048,
|
| 5 |
+
"transformers_version": "4.51.3"
|
| 6 |
+
}
|
checkpoint-5751/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bb3d1bd95c53800ef1cb04bf0b4dca4678ae38f01058094213011b8ab1df2db9
|
| 3 |
+
size 1192845752
|
checkpoint-5751/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1f80e3bfb0a586f776698e79ea6153229419cd4d77b77c70d4fa9d4c9ead70a5
|
| 3 |
+
size 622337116
|
checkpoint-5751/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14244
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1ff264f99d31b522cc7e2a4eac9d38606d0c58a34c0adc74d71e0ca8b371dc36
|
| 3 |
size 14244
|
checkpoint-5751/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1064
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:903ca14f005e51a1e28984de583841d853b88748f2e6d3a88d37904a28596d8e
|
| 3 |
size 1064
|
checkpoint-5751/trainer_state.json
CHANGED
|
@@ -11,87 +11,87 @@
|
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
"epoch": 0.0869414014953921,
|
| 14 |
-
"grad_norm":
|
| 15 |
-
"learning_rate": 0.
|
| 16 |
-
"loss":
|
| 17 |
"step": 500
|
| 18 |
},
|
| 19 |
{
|
| 20 |
"epoch": 0.1738828029907842,
|
| 21 |
-
"grad_norm":
|
| 22 |
-
"learning_rate": 0.
|
| 23 |
-
"loss": 0.
|
| 24 |
"step": 1000
|
| 25 |
},
|
| 26 |
{
|
| 27 |
"epoch": 0.2608242044861763,
|
| 28 |
-
"grad_norm":
|
| 29 |
-
"learning_rate": 0.
|
| 30 |
-
"loss": 0.
|
| 31 |
"step": 1500
|
| 32 |
},
|
| 33 |
{
|
| 34 |
"epoch": 0.3477656059815684,
|
| 35 |
-
"grad_norm":
|
| 36 |
-
"learning_rate": 0.
|
| 37 |
-
"loss": 0.
|
| 38 |
"step": 2000
|
| 39 |
},
|
| 40 |
{
|
| 41 |
"epoch": 0.4347070074769605,
|
| 42 |
-
"grad_norm":
|
| 43 |
-
"learning_rate": 0.
|
| 44 |
-
"loss": 0.
|
| 45 |
"step": 2500
|
| 46 |
},
|
| 47 |
{
|
| 48 |
"epoch": 0.5216484089723527,
|
| 49 |
-
"grad_norm":
|
| 50 |
-
"learning_rate": 0.
|
| 51 |
-
"loss": 0.
|
| 52 |
"step": 3000
|
| 53 |
},
|
| 54 |
{
|
| 55 |
"epoch": 0.6085898104677447,
|
| 56 |
-
"grad_norm":
|
| 57 |
-
"learning_rate": 0.
|
| 58 |
-
"loss": 0.
|
| 59 |
"step": 3500
|
| 60 |
},
|
| 61 |
{
|
| 62 |
"epoch": 0.6955312119631368,
|
| 63 |
-
"grad_norm":
|
| 64 |
-
"learning_rate": 0.
|
| 65 |
-
"loss": 0.
|
| 66 |
"step": 4000
|
| 67 |
},
|
| 68 |
{
|
| 69 |
"epoch": 0.7824726134585289,
|
| 70 |
-
"grad_norm":
|
| 71 |
-
"learning_rate": 0.
|
| 72 |
-
"loss": 0.
|
| 73 |
"step": 4500
|
| 74 |
},
|
| 75 |
{
|
| 76 |
"epoch": 0.869414014953921,
|
| 77 |
-
"grad_norm":
|
| 78 |
-
"learning_rate": 0.
|
| 79 |
-
"loss": 0.
|
| 80 |
"step": 5000
|
| 81 |
},
|
| 82 |
{
|
| 83 |
"epoch": 0.9563554164493132,
|
| 84 |
-
"grad_norm":
|
| 85 |
-
"learning_rate": 0.
|
| 86 |
-
"loss": 0.
|
| 87 |
"step": 5500
|
| 88 |
},
|
| 89 |
{
|
| 90 |
"epoch": 1.0,
|
| 91 |
-
"eval_loss":
|
| 92 |
-
"eval_runtime":
|
| 93 |
-
"eval_samples_per_second":
|
| 94 |
-
"eval_steps_per_second":
|
| 95 |
"step": 5751
|
| 96 |
}
|
| 97 |
],
|
|
@@ -112,7 +112,7 @@
|
|
| 112 |
"attributes": {}
|
| 113 |
}
|
| 114 |
},
|
| 115 |
-
"total_flos": 3.
|
| 116 |
"train_batch_size": 4,
|
| 117 |
"trial_name": null,
|
| 118 |
"trial_params": null
|
|
|
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
"epoch": 0.0869414014953921,
|
| 14 |
+
"grad_norm": NaN,
|
| 15 |
+
"learning_rate": 0.00019421549875383993,
|
| 16 |
+
"loss": 4.2823,
|
| 17 |
"step": 500
|
| 18 |
},
|
| 19 |
{
|
| 20 |
"epoch": 0.1738828029907842,
|
| 21 |
+
"grad_norm": NaN,
|
| 22 |
+
"learning_rate": 0.0001884194053208138,
|
| 23 |
+
"loss": 0.0,
|
| 24 |
"step": 1000
|
| 25 |
},
|
| 26 |
{
|
| 27 |
"epoch": 0.2608242044861763,
|
| 28 |
+
"grad_norm": NaN,
|
| 29 |
+
"learning_rate": 0.00018262331188778763,
|
| 30 |
+
"loss": 0.0,
|
| 31 |
"step": 1500
|
| 32 |
},
|
| 33 |
{
|
| 34 |
"epoch": 0.3477656059815684,
|
| 35 |
+
"grad_norm": NaN,
|
| 36 |
+
"learning_rate": 0.0001768272184547615,
|
| 37 |
+
"loss": 0.0,
|
| 38 |
"step": 2000
|
| 39 |
},
|
| 40 |
{
|
| 41 |
"epoch": 0.4347070074769605,
|
| 42 |
+
"grad_norm": NaN,
|
| 43 |
+
"learning_rate": 0.00017103112502173535,
|
| 44 |
+
"loss": 0.0,
|
| 45 |
"step": 2500
|
| 46 |
},
|
| 47 |
{
|
| 48 |
"epoch": 0.5216484089723527,
|
| 49 |
+
"grad_norm": NaN,
|
| 50 |
+
"learning_rate": 0.0001652350315887092,
|
| 51 |
+
"loss": 0.0,
|
| 52 |
"step": 3000
|
| 53 |
},
|
| 54 |
{
|
| 55 |
"epoch": 0.6085898104677447,
|
| 56 |
+
"grad_norm": NaN,
|
| 57 |
+
"learning_rate": 0.00015943893815568308,
|
| 58 |
+
"loss": 0.0,
|
| 59 |
"step": 3500
|
| 60 |
},
|
| 61 |
{
|
| 62 |
"epoch": 0.6955312119631368,
|
| 63 |
+
"grad_norm": NaN,
|
| 64 |
+
"learning_rate": 0.00015364284472265693,
|
| 65 |
+
"loss": 0.0,
|
| 66 |
"step": 4000
|
| 67 |
},
|
| 68 |
{
|
| 69 |
"epoch": 0.7824726134585289,
|
| 70 |
+
"grad_norm": NaN,
|
| 71 |
+
"learning_rate": 0.0001478467512896308,
|
| 72 |
+
"loss": 0.0,
|
| 73 |
"step": 4500
|
| 74 |
},
|
| 75 |
{
|
| 76 |
"epoch": 0.869414014953921,
|
| 77 |
+
"grad_norm": NaN,
|
| 78 |
+
"learning_rate": 0.00014205065785660465,
|
| 79 |
+
"loss": 0.0,
|
| 80 |
"step": 5000
|
| 81 |
},
|
| 82 |
{
|
| 83 |
"epoch": 0.9563554164493132,
|
| 84 |
+
"grad_norm": NaN,
|
| 85 |
+
"learning_rate": 0.0001362545644235785,
|
| 86 |
+
"loss": 0.0,
|
| 87 |
"step": 5500
|
| 88 |
},
|
| 89 |
{
|
| 90 |
"epoch": 1.0,
|
| 91 |
+
"eval_loss": NaN,
|
| 92 |
+
"eval_runtime": 152.5081,
|
| 93 |
+
"eval_samples_per_second": 16.76,
|
| 94 |
+
"eval_steps_per_second": 4.19,
|
| 95 |
"step": 5751
|
| 96 |
}
|
| 97 |
],
|
|
|
|
| 112 |
"attributes": {}
|
| 113 |
}
|
| 114 |
},
|
| 115 |
+
"total_flos": 3.115004127608832e+16,
|
| 116 |
"train_batch_size": 4,
|
| 117 |
"trial_name": null,
|
| 118 |
"trial_params": null
|
checkpoint-5751/training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5240
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:99fccf4927862cc4fd8283f3958f1ef8528c86594c8ff39225b274adde4416df
|
| 3 |
size 5240
|