Upload folder using huggingface_hub
Browse files- checkpoints-v3/checkpoint-7168/eval_state.json +0 -0
- checkpoints-v3/checkpoint-7168/model.safetensors +3 -0
- checkpoints-v3/checkpoint-7168/optimizer.pt +3 -0
- checkpoints-v3/checkpoint-7168/rng_state.pth +3 -0
- checkpoints-v3/checkpoint-7168/scaler.pt +3 -0
- checkpoints-v3/checkpoint-7168/scheduler.pt +3 -0
- checkpoints-v3/checkpoint-7168/trainer_state.json +573 -0
- checkpoints-v3/checkpoint-7168/training_args.bin +3 -0
checkpoints-v3/checkpoint-7168/eval_state.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoints-v3/checkpoint-7168/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6f7daf11bb4eaab952771d897197a4c2cf99e395c51912e1a734b70b8761a043
|
| 3 |
+
size 37402680
|
checkpoints-v3/checkpoint-7168/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c895e6a2893c3a519125ef0880ccd3b676b305dfca932aedac9bc49b9c54e7d9
|
| 3 |
+
size 512267
|
checkpoints-v3/checkpoint-7168/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:01c01c678fa29a9262064ade7c90c5c12c8ca1b0c16a002081bbaef811e78e00
|
| 3 |
+
size 14645
|
checkpoints-v3/checkpoint-7168/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d595dc04756955b718dbe40e12e3b42e9a74ec09bbdeec39a22714665de3cd13
|
| 3 |
+
size 1383
|
checkpoints-v3/checkpoint-7168/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1accfb8c9e4435c988a0a31d91d0a223e4cb602232ad6ce46b39c8faafb8ab9d
|
| 3 |
+
size 1465
|
checkpoints-v3/checkpoint-7168/trainer_state.json
ADDED
|
@@ -0,0 +1,573 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.07450407704020913,
|
| 6 |
+
"eval_steps": 1024,
|
| 7 |
+
"global_step": 7168,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.002660859894293183,
|
| 14 |
+
"grad_norm": 1.2474960088729858,
|
| 15 |
+
"learning_rate": 4.416349151368202e-06,
|
| 16 |
+
"loss": 10.7959,
|
| 17 |
+
"step": 256
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.005321719788586366,
|
| 21 |
+
"grad_norm": 0.9909601807594299,
|
| 22 |
+
"learning_rate": 8.85001731901628e-06,
|
| 23 |
+
"loss": 10.0541,
|
| 24 |
+
"step": 512
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.00798257968287955,
|
| 28 |
+
"grad_norm": 0.8707857728004456,
|
| 29 |
+
"learning_rate": 1.3283685486664357e-05,
|
| 30 |
+
"loss": 9.0128,
|
| 31 |
+
"step": 768
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.010643439577172733,
|
| 35 |
+
"grad_norm": 0.8646675944328308,
|
| 36 |
+
"learning_rate": 1.7717353654312436e-05,
|
| 37 |
+
"loss": 8.0258,
|
| 38 |
+
"step": 1024
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.010643439577172733,
|
| 42 |
+
"eval_acr_loss": 0.010622835965477861,
|
| 43 |
+
"eval_across_var": 0.012448028806829825,
|
| 44 |
+
"eval_bleu": 0.09796718659972453,
|
| 45 |
+
"eval_ce_loss": 7.011146575212479,
|
| 46 |
+
"eval_cos_loss": 0.9561333861202002,
|
| 47 |
+
"eval_cov": 0.06968498229980469,
|
| 48 |
+
"eval_cov_loss": 0.007897357034380548,
|
| 49 |
+
"eval_glb_loss": 0.4064090773463249,
|
| 50 |
+
"eval_global_kurtosis": 3.1536246612668037,
|
| 51 |
+
"eval_global_mean": -0.0043766796588897705,
|
| 52 |
+
"eval_global_var": 0.26236724853515625,
|
| 53 |
+
"eval_krt_loss": 0.02414830235647969,
|
| 54 |
+
"eval_loss": 7.440777659416199,
|
| 55 |
+
"eval_mean_loss": 2.432924324580199e-05,
|
| 56 |
+
"eval_mse_loss": 1.9190465211868286,
|
| 57 |
+
"eval_per_loss": 0.3546135723590851,
|
| 58 |
+
"eval_per_var": 0.25460052490234375,
|
| 59 |
+
"eval_within_var": 0.2497538859024644,
|
| 60 |
+
"eval_wth_loss": 0.42283743154257536,
|
| 61 |
+
"step": 1024
|
| 62 |
+
},
|
| 63 |
+
{
|
| 64 |
+
"epoch": 0.010643439577172733,
|
| 65 |
+
"eval_acr_loss": 0.010622835965477861,
|
| 66 |
+
"eval_across_var": 0.012448028806829825,
|
| 67 |
+
"eval_bleu": 0.09796718659972453,
|
| 68 |
+
"eval_ce_loss": 7.011146575212479,
|
| 69 |
+
"eval_cos_loss": 0.9561333861202002,
|
| 70 |
+
"eval_cov": 0.06968498229980469,
|
| 71 |
+
"eval_cov_loss": 0.007897357034380548,
|
| 72 |
+
"eval_glb_loss": 0.4064090773463249,
|
| 73 |
+
"eval_global_kurtosis": 3.1536246612668037,
|
| 74 |
+
"eval_global_mean": -0.0043766796588897705,
|
| 75 |
+
"eval_global_var": 0.26236724853515625,
|
| 76 |
+
"eval_krt_loss": 0.02414830235647969,
|
| 77 |
+
"eval_loss": 7.440777659416199,
|
| 78 |
+
"eval_mean_loss": 2.432924324580199e-05,
|
| 79 |
+
"eval_mse_loss": 1.9190465211868286,
|
| 80 |
+
"eval_per_loss": 0.3546135723590851,
|
| 81 |
+
"eval_per_var": 0.25460052490234375,
|
| 82 |
+
"eval_runtime": 10.2924,
|
| 83 |
+
"eval_samples_per_second": 194.318,
|
| 84 |
+
"eval_steps_per_second": 3.109,
|
| 85 |
+
"eval_within_var": 0.2497538859024644,
|
| 86 |
+
"eval_wth_loss": 0.42283743154257536,
|
| 87 |
+
"step": 1024
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.013304299471465915,
|
| 91 |
+
"grad_norm": 0.8640124797821045,
|
| 92 |
+
"learning_rate": 2.2151021821960514e-05,
|
| 93 |
+
"loss": 7.0717,
|
| 94 |
+
"step": 1280
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.0159651593657591,
|
| 98 |
+
"grad_norm": 0.7730758190155029,
|
| 99 |
+
"learning_rate": 2.6584689989608592e-05,
|
| 100 |
+
"loss": 6.1186,
|
| 101 |
+
"step": 1536
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.018626019260052282,
|
| 105 |
+
"grad_norm": 0.6963288187980652,
|
| 106 |
+
"learning_rate": 3.1018358157256674e-05,
|
| 107 |
+
"loss": 5.2034,
|
| 108 |
+
"step": 1792
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.021286879154345465,
|
| 112 |
+
"grad_norm": 0.5616676211357117,
|
| 113 |
+
"learning_rate": 3.5452026324904745e-05,
|
| 114 |
+
"loss": 4.3736,
|
| 115 |
+
"step": 2048
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.021286879154345465,
|
| 119 |
+
"eval_acr_loss": 0.012247161677805707,
|
| 120 |
+
"eval_across_var": 0.012088194896932691,
|
| 121 |
+
"eval_bleu": 0.3414767015551533,
|
| 122 |
+
"eval_ce_loss": 3.459592819213867,
|
| 123 |
+
"eval_cos_loss": 0.9247305598109961,
|
| 124 |
+
"eval_cov": 0.0738067626953125,
|
| 125 |
+
"eval_cov_loss": 0.008805307501461357,
|
| 126 |
+
"eval_glb_loss": 0.354451559484005,
|
| 127 |
+
"eval_global_kurtosis": 3.0689163729548454,
|
| 128 |
+
"eval_global_mean": -0.004139065742492676,
|
| 129 |
+
"eval_global_var": 0.30448150634765625,
|
| 130 |
+
"eval_krt_loss": 0.005096593113194103,
|
| 131 |
+
"eval_loss": 3.872511200606823,
|
| 132 |
+
"eval_mean_loss": 2.0987964205687604e-05,
|
| 133 |
+
"eval_mse_loss": 1.9079551436007023,
|
| 134 |
+
"eval_per_loss": 0.3076172471046448,
|
| 135 |
+
"eval_per_var": 0.29547882080078125,
|
| 136 |
+
"eval_within_var": 0.29124921560287476,
|
| 137 |
+
"eval_wth_loss": 0.3706007469445467,
|
| 138 |
+
"step": 2048
|
| 139 |
+
},
|
| 140 |
+
{
|
| 141 |
+
"epoch": 0.021286879154345465,
|
| 142 |
+
"eval_acr_loss": 0.012247161677805707,
|
| 143 |
+
"eval_across_var": 0.012088194896932691,
|
| 144 |
+
"eval_bleu": 0.3414767015551533,
|
| 145 |
+
"eval_ce_loss": 3.459592819213867,
|
| 146 |
+
"eval_cos_loss": 0.9247305598109961,
|
| 147 |
+
"eval_cov": 0.0738067626953125,
|
| 148 |
+
"eval_cov_loss": 0.008805307501461357,
|
| 149 |
+
"eval_glb_loss": 0.354451559484005,
|
| 150 |
+
"eval_global_kurtosis": 3.0689163729548454,
|
| 151 |
+
"eval_global_mean": -0.004139065742492676,
|
| 152 |
+
"eval_global_var": 0.30448150634765625,
|
| 153 |
+
"eval_krt_loss": 0.005096593113194103,
|
| 154 |
+
"eval_loss": 3.872511200606823,
|
| 155 |
+
"eval_mean_loss": 2.0987964205687604e-05,
|
| 156 |
+
"eval_mse_loss": 1.9079551436007023,
|
| 157 |
+
"eval_per_loss": 0.3076172471046448,
|
| 158 |
+
"eval_per_var": 0.29547882080078125,
|
| 159 |
+
"eval_runtime": 9.9029,
|
| 160 |
+
"eval_samples_per_second": 201.962,
|
| 161 |
+
"eval_steps_per_second": 3.231,
|
| 162 |
+
"eval_within_var": 0.29124921560287476,
|
| 163 |
+
"eval_wth_loss": 0.3706007469445467,
|
| 164 |
+
"step": 2048
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.023947739048638648,
|
| 168 |
+
"grad_norm": 0.44369781017303467,
|
| 169 |
+
"learning_rate": 3.988569449255283e-05,
|
| 170 |
+
"loss": 3.6681,
|
| 171 |
+
"step": 2304
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.02660859894293183,
|
| 175 |
+
"grad_norm": 0.3687000274658203,
|
| 176 |
+
"learning_rate": 4.43193626602009e-05,
|
| 177 |
+
"loss": 3.0886,
|
| 178 |
+
"step": 2560
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.029269458837225013,
|
| 182 |
+
"grad_norm": 0.3568866550922394,
|
| 183 |
+
"learning_rate": 4.875303082784898e-05,
|
| 184 |
+
"loss": 2.6049,
|
| 185 |
+
"step": 2816
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.0319303187315182,
|
| 189 |
+
"grad_norm": 0.2986361086368561,
|
| 190 |
+
"learning_rate": 4.9999520413849384e-05,
|
| 191 |
+
"loss": 2.2063,
|
| 192 |
+
"step": 3072
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.0319303187315182,
|
| 196 |
+
"eval_acr_loss": 0.011944463331019506,
|
| 197 |
+
"eval_across_var": 0.025048962590517476,
|
| 198 |
+
"eval_bleu": 0.5783938497071468,
|
| 199 |
+
"eval_ce_loss": 1.5567151941359043,
|
| 200 |
+
"eval_cos_loss": 0.8373732026666403,
|
| 201 |
+
"eval_cov": 0.10790634155273438,
|
| 202 |
+
"eval_cov_loss": 0.01786720016389154,
|
| 203 |
+
"eval_glb_loss": 0.08421005308628082,
|
| 204 |
+
"eval_global_kurtosis": 3.042153775691986,
|
| 205 |
+
"eval_global_mean": -0.0013459473848342896,
|
| 206 |
+
"eval_global_var": 0.60980224609375,
|
| 207 |
+
"eval_krt_loss": 0.0020023469523948734,
|
| 208 |
+
"eval_loss": 1.9125033244490623,
|
| 209 |
+
"eval_mean_loss": 9.189085614202952e-06,
|
| 210 |
+
"eval_mse_loss": 1.798950683325529,
|
| 211 |
+
"eval_per_loss": 0.06572123290970922,
|
| 212 |
+
"eval_per_var": 0.59381103515625,
|
| 213 |
+
"eval_within_var": 0.5743193719536066,
|
| 214 |
+
"eval_wth_loss": 0.10620259935967624,
|
| 215 |
+
"step": 3072
|
| 216 |
+
},
|
| 217 |
+
{
|
| 218 |
+
"epoch": 0.0319303187315182,
|
| 219 |
+
"eval_acr_loss": 0.011944463331019506,
|
| 220 |
+
"eval_across_var": 0.025048962590517476,
|
| 221 |
+
"eval_bleu": 0.5783938497071468,
|
| 222 |
+
"eval_ce_loss": 1.5567151941359043,
|
| 223 |
+
"eval_cos_loss": 0.8373732026666403,
|
| 224 |
+
"eval_cov": 0.10790634155273438,
|
| 225 |
+
"eval_cov_loss": 0.01786720016389154,
|
| 226 |
+
"eval_glb_loss": 0.08421005308628082,
|
| 227 |
+
"eval_global_kurtosis": 3.042153775691986,
|
| 228 |
+
"eval_global_mean": -0.0013459473848342896,
|
| 229 |
+
"eval_global_var": 0.60980224609375,
|
| 230 |
+
"eval_krt_loss": 0.0020023469523948734,
|
| 231 |
+
"eval_loss": 1.9125033244490623,
|
| 232 |
+
"eval_mean_loss": 9.189085614202952e-06,
|
| 233 |
+
"eval_mse_loss": 1.798950683325529,
|
| 234 |
+
"eval_per_loss": 0.06572123290970922,
|
| 235 |
+
"eval_per_var": 0.59381103515625,
|
| 236 |
+
"eval_runtime": 10.4973,
|
| 237 |
+
"eval_samples_per_second": 190.525,
|
| 238 |
+
"eval_steps_per_second": 3.048,
|
| 239 |
+
"eval_within_var": 0.5743193719536066,
|
| 240 |
+
"eval_wth_loss": 0.10620259935967624,
|
| 241 |
+
"step": 3072
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.03459117862581138,
|
| 245 |
+
"grad_norm": 0.270656943321228,
|
| 246 |
+
"learning_rate": 4.9997257606389056e-05,
|
| 247 |
+
"loss": 1.8881,
|
| 248 |
+
"step": 3328
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.037252038520104565,
|
| 252 |
+
"grad_norm": 0.24188651144504547,
|
| 253 |
+
"learning_rate": 4.999313831167736e-05,
|
| 254 |
+
"loss": 1.6388,
|
| 255 |
+
"step": 3584
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.03991289841439775,
|
| 259 |
+
"grad_norm": 0.2294900268316269,
|
| 260 |
+
"learning_rate": 4.998716283564454e-05,
|
| 261 |
+
"loss": 1.4382,
|
| 262 |
+
"step": 3840
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.04257375830869093,
|
| 266 |
+
"grad_norm": 0.20773501694202423,
|
| 267 |
+
"learning_rate": 4.99793316220751e-05,
|
| 268 |
+
"loss": 1.2713,
|
| 269 |
+
"step": 4096
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.04257375830869093,
|
| 273 |
+
"eval_acr_loss": 0.011318061951897107,
|
| 274 |
+
"eval_across_var": 0.037467821151949465,
|
| 275 |
+
"eval_bleu": 0.7398917090811331,
|
| 276 |
+
"eval_ce_loss": 0.8147697541862726,
|
| 277 |
+
"eval_cos_loss": 0.739902313798666,
|
| 278 |
+
"eval_cov": 0.09920120239257812,
|
| 279 |
+
"eval_cov_loss": 0.015144521807087585,
|
| 280 |
+
"eval_glb_loss": 0.0026292089896742254,
|
| 281 |
+
"eval_global_kurtosis": 3.0431209057569504,
|
| 282 |
+
"eval_global_mean": 0.0002828165888786316,
|
| 283 |
+
"eval_global_var": 0.849639892578125,
|
| 284 |
+
"eval_krt_loss": 0.002429395680081825,
|
| 285 |
+
"eval_loss": 1.1247494276612997,
|
| 286 |
+
"eval_mean_loss": 1.0841616662204956e-05,
|
| 287 |
+
"eval_mse_loss": 1.6548683494329453,
|
| 288 |
+
"eval_per_loss": 0.0006620931362704141,
|
| 289 |
+
"eval_per_var": 0.826202392578125,
|
| 290 |
+
"eval_within_var": 0.8049256391823292,
|
| 291 |
+
"eval_wth_loss": 0.009282346058171242,
|
| 292 |
+
"step": 4096
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 0.04257375830869093,
|
| 296 |
+
"eval_acr_loss": 0.011318061951897107,
|
| 297 |
+
"eval_across_var": 0.037467821151949465,
|
| 298 |
+
"eval_bleu": 0.7398917090811331,
|
| 299 |
+
"eval_ce_loss": 0.8147697541862726,
|
| 300 |
+
"eval_cos_loss": 0.739902313798666,
|
| 301 |
+
"eval_cov": 0.09920120239257812,
|
| 302 |
+
"eval_cov_loss": 0.015144521807087585,
|
| 303 |
+
"eval_glb_loss": 0.0026292089896742254,
|
| 304 |
+
"eval_global_kurtosis": 3.0431209057569504,
|
| 305 |
+
"eval_global_mean": 0.0002828165888786316,
|
| 306 |
+
"eval_global_var": 0.849639892578125,
|
| 307 |
+
"eval_krt_loss": 0.002429395680081825,
|
| 308 |
+
"eval_loss": 1.1247494276612997,
|
| 309 |
+
"eval_mean_loss": 1.0841616662204956e-05,
|
| 310 |
+
"eval_mse_loss": 1.6548683494329453,
|
| 311 |
+
"eval_per_loss": 0.0006620931362704141,
|
| 312 |
+
"eval_per_var": 0.826202392578125,
|
| 313 |
+
"eval_runtime": 10.0939,
|
| 314 |
+
"eval_samples_per_second": 198.14,
|
| 315 |
+
"eval_steps_per_second": 3.17,
|
| 316 |
+
"eval_within_var": 0.8049256391823292,
|
| 317 |
+
"eval_wth_loss": 0.009282346058171242,
|
| 318 |
+
"step": 4096
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.04523461820298411,
|
| 322 |
+
"grad_norm": 0.1941945105791092,
|
| 323 |
+
"learning_rate": 4.996964525257477e-05,
|
| 324 |
+
"loss": 1.1364,
|
| 325 |
+
"step": 4352
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.047895478097277296,
|
| 329 |
+
"grad_norm": 0.17706365883350372,
|
| 330 |
+
"learning_rate": 4.995810444652731e-05,
|
| 331 |
+
"loss": 1.0202,
|
| 332 |
+
"step": 4608
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.05055633799157048,
|
| 336 |
+
"grad_norm": 0.17764592170715332,
|
| 337 |
+
"learning_rate": 4.994471006104112e-05,
|
| 338 |
+
"loss": 0.9256,
|
| 339 |
+
"step": 4864
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.05321719788586366,
|
| 343 |
+
"grad_norm": 0.1597519963979721,
|
| 344 |
+
"learning_rate": 4.992946309088557e-05,
|
| 345 |
+
"loss": 0.8433,
|
| 346 |
+
"step": 5120
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.05321719788586366,
|
| 350 |
+
"eval_acr_loss": 0.010796478512929752,
|
| 351 |
+
"eval_across_var": 0.0437286015949212,
|
| 352 |
+
"eval_bleu": 0.8359253777154618,
|
| 353 |
+
"eval_ce_loss": 0.4896330190822482,
|
| 354 |
+
"eval_cos_loss": 0.6560099385678768,
|
| 355 |
+
"eval_cov": 0.08585166931152344,
|
| 356 |
+
"eval_cov_loss": 0.011597162316320464,
|
| 357 |
+
"eval_glb_loss": 0.0,
|
| 358 |
+
"eval_global_kurtosis": 3.051339641213417,
|
| 359 |
+
"eval_global_mean": 0.0004043206572532654,
|
| 360 |
+
"eval_global_var": 0.9364166259765625,
|
| 361 |
+
"eval_krt_loss": 0.0034133701161636054,
|
| 362 |
+
"eval_loss": 0.7659010197967291,
|
| 363 |
+
"eval_mean_loss": 1.1450480416286268e-05,
|
| 364 |
+
"eval_mse_loss": 1.5238465368747711,
|
| 365 |
+
"eval_per_loss": 0.0,
|
| 366 |
+
"eval_per_var": 0.9102935791015625,
|
| 367 |
+
"eval_within_var": 0.8955719340592623,
|
| 368 |
+
"eval_wth_loss": 0.00021194279955238926,
|
| 369 |
+
"step": 5120
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 0.05321719788586366,
|
| 373 |
+
"eval_acr_loss": 0.010796478512929752,
|
| 374 |
+
"eval_across_var": 0.0437286015949212,
|
| 375 |
+
"eval_bleu": 0.8359253777154618,
|
| 376 |
+
"eval_ce_loss": 0.4896330190822482,
|
| 377 |
+
"eval_cos_loss": 0.6560099385678768,
|
| 378 |
+
"eval_cov": 0.08585166931152344,
|
| 379 |
+
"eval_cov_loss": 0.011597162316320464,
|
| 380 |
+
"eval_glb_loss": 0.0,
|
| 381 |
+
"eval_global_kurtosis": 3.051339641213417,
|
| 382 |
+
"eval_global_mean": 0.0004043206572532654,
|
| 383 |
+
"eval_global_var": 0.9364166259765625,
|
| 384 |
+
"eval_krt_loss": 0.0034133701161636054,
|
| 385 |
+
"eval_loss": 0.7659010197967291,
|
| 386 |
+
"eval_mean_loss": 1.1450480416286268e-05,
|
| 387 |
+
"eval_mse_loss": 1.5238465368747711,
|
| 388 |
+
"eval_per_loss": 0.0,
|
| 389 |
+
"eval_per_var": 0.9102935791015625,
|
| 390 |
+
"eval_runtime": 10.0333,
|
| 391 |
+
"eval_samples_per_second": 199.336,
|
| 392 |
+
"eval_steps_per_second": 3.189,
|
| 393 |
+
"eval_within_var": 0.8955719340592623,
|
| 394 |
+
"eval_wth_loss": 0.00021194279955238926,
|
| 395 |
+
"step": 5120
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.055878057780156844,
|
| 399 |
+
"grad_norm": 0.15128253400325775,
|
| 400 |
+
"learning_rate": 4.991236466841708e-05,
|
| 401 |
+
"loss": 0.7748,
|
| 402 |
+
"step": 5376
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.058538917674450026,
|
| 406 |
+
"grad_norm": 0.15075387060642242,
|
| 407 |
+
"learning_rate": 4.989341606349509e-05,
|
| 408 |
+
"loss": 0.7149,
|
| 409 |
+
"step": 5632
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.06119977756874321,
|
| 413 |
+
"grad_norm": 0.13722559809684753,
|
| 414 |
+
"learning_rate": 4.987261868338772e-05,
|
| 415 |
+
"loss": 0.6633,
|
| 416 |
+
"step": 5888
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.0638606374630364,
|
| 420 |
+
"grad_norm": 0.14299507439136505,
|
| 421 |
+
"learning_rate": 4.9849974072667235e-05,
|
| 422 |
+
"loss": 0.6168,
|
| 423 |
+
"step": 6144
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.0638606374630364,
|
| 427 |
+
"eval_acr_loss": 0.010568196172243915,
|
| 428 |
+
"eval_across_var": 0.050391704426147044,
|
| 429 |
+
"eval_bleu": 0.8864417334039504,
|
| 430 |
+
"eval_ce_loss": 0.3192982799373567,
|
| 431 |
+
"eval_cos_loss": 0.5848112031817436,
|
| 432 |
+
"eval_cov": 0.08610343933105469,
|
| 433 |
+
"eval_cov_loss": 0.011645367194432765,
|
| 434 |
+
"eval_glb_loss": 0.0,
|
| 435 |
+
"eval_global_kurtosis": 3.057781808078289,
|
| 436 |
+
"eval_global_mean": 0.00010520219802856445,
|
| 437 |
+
"eval_global_var": 1.05322265625,
|
| 438 |
+
"eval_krt_loss": 0.00413643000592856,
|
| 439 |
+
"eval_loss": 0.5672316299751401,
|
| 440 |
+
"eval_mean_loss": 1.1898590268621945e-05,
|
| 441 |
+
"eval_mse_loss": 1.4093649201095104,
|
| 442 |
+
"eval_per_loss": 0.0,
|
| 443 |
+
"eval_per_var": 1.0248565673828125,
|
| 444 |
+
"eval_within_var": 1.0096650514751673,
|
| 445 |
+
"eval_wth_loss": 0.0,
|
| 446 |
+
"step": 6144
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 0.0638606374630364,
|
| 450 |
+
"eval_acr_loss": 0.010568196172243915,
|
| 451 |
+
"eval_across_var": 0.050391704426147044,
|
| 452 |
+
"eval_bleu": 0.8864417334039504,
|
| 453 |
+
"eval_ce_loss": 0.3192982799373567,
|
| 454 |
+
"eval_cos_loss": 0.5848112031817436,
|
| 455 |
+
"eval_cov": 0.08610343933105469,
|
| 456 |
+
"eval_cov_loss": 0.011645367194432765,
|
| 457 |
+
"eval_glb_loss": 0.0,
|
| 458 |
+
"eval_global_kurtosis": 3.057781808078289,
|
| 459 |
+
"eval_global_mean": 0.00010520219802856445,
|
| 460 |
+
"eval_global_var": 1.05322265625,
|
| 461 |
+
"eval_krt_loss": 0.00413643000592856,
|
| 462 |
+
"eval_loss": 0.5672316299751401,
|
| 463 |
+
"eval_mean_loss": 1.1898590268621945e-05,
|
| 464 |
+
"eval_mse_loss": 1.4093649201095104,
|
| 465 |
+
"eval_per_loss": 0.0,
|
| 466 |
+
"eval_per_var": 1.0248565673828125,
|
| 467 |
+
"eval_runtime": 10.495,
|
| 468 |
+
"eval_samples_per_second": 190.567,
|
| 469 |
+
"eval_steps_per_second": 3.049,
|
| 470 |
+
"eval_within_var": 1.0096650514751673,
|
| 471 |
+
"eval_wth_loss": 0.0,
|
| 472 |
+
"step": 6144
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.06652149735732958,
|
| 476 |
+
"grad_norm": 0.13175231218338013,
|
| 477 |
+
"learning_rate": 4.9825483913095364e-05,
|
| 478 |
+
"loss": 0.5727,
|
| 479 |
+
"step": 6400
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.06918235725162276,
|
| 483 |
+
"grad_norm": 0.130602166056633,
|
| 484 |
+
"learning_rate": 4.979915002349838e-05,
|
| 485 |
+
"loss": 0.5411,
|
| 486 |
+
"step": 6656
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 0.07184321714591595,
|
| 490 |
+
"grad_norm": 0.12843571603298187,
|
| 491 |
+
"learning_rate": 4.977097435963204e-05,
|
| 492 |
+
"loss": 0.5082,
|
| 493 |
+
"step": 6912
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.07450407704020913,
|
| 497 |
+
"grad_norm": 0.1221570074558258,
|
| 498 |
+
"learning_rate": 4.974095901403632e-05,
|
| 499 |
+
"loss": 0.4775,
|
| 500 |
+
"step": 7168
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.07450407704020913,
|
| 504 |
+
"eval_acr_loss": 0.01032613120332826,
|
| 505 |
+
"eval_across_var": 0.055548187578096986,
|
| 506 |
+
"eval_bleu": 0.917825685067053,
|
| 507 |
+
"eval_ce_loss": 0.22222592495381832,
|
| 508 |
+
"eval_cos_loss": 0.5258319452404976,
|
| 509 |
+
"eval_cov": 0.0838165283203125,
|
| 510 |
+
"eval_cov_loss": 0.01109178303158842,
|
| 511 |
+
"eval_glb_loss": 0.0011626811420910599,
|
| 512 |
+
"eval_global_kurtosis": 3.0560965314507484,
|
| 513 |
+
"eval_global_mean": -0.00032412633299827576,
|
| 514 |
+
"eval_global_var": 1.1317138671875,
|
| 515 |
+
"eval_krt_loss": 0.003975647037577801,
|
| 516 |
+
"eval_loss": 0.4466686090454459,
|
| 517 |
+
"eval_mean_loss": 1.2274138283974168e-05,
|
| 518 |
+
"eval_mse_loss": 1.3142655715346336,
|
| 519 |
+
"eval_per_loss": 0.0,
|
| 520 |
+
"eval_per_var": 1.1024169921875,
|
| 521 |
+
"eval_within_var": 1.088273286819458,
|
| 522 |
+
"eval_wth_loss": 4.5452433568016204e-05,
|
| 523 |
+
"step": 7168
|
| 524 |
+
},
|
| 525 |
+
{
|
| 526 |
+
"epoch": 0.07450407704020913,
|
| 527 |
+
"eval_acr_loss": 0.01032613120332826,
|
| 528 |
+
"eval_across_var": 0.055548187578096986,
|
| 529 |
+
"eval_bleu": 0.917825685067053,
|
| 530 |
+
"eval_ce_loss": 0.22222592495381832,
|
| 531 |
+
"eval_cos_loss": 0.5258319452404976,
|
| 532 |
+
"eval_cov": 0.0838165283203125,
|
| 533 |
+
"eval_cov_loss": 0.01109178303158842,
|
| 534 |
+
"eval_glb_loss": 0.0011626811420910599,
|
| 535 |
+
"eval_global_kurtosis": 3.0560965314507484,
|
| 536 |
+
"eval_global_mean": -0.00032412633299827576,
|
| 537 |
+
"eval_global_var": 1.1317138671875,
|
| 538 |
+
"eval_krt_loss": 0.003975647037577801,
|
| 539 |
+
"eval_loss": 0.4466686090454459,
|
| 540 |
+
"eval_mean_loss": 1.2274138283974168e-05,
|
| 541 |
+
"eval_mse_loss": 1.3142655715346336,
|
| 542 |
+
"eval_per_loss": 0.0,
|
| 543 |
+
"eval_per_var": 1.1024169921875,
|
| 544 |
+
"eval_runtime": 10.2975,
|
| 545 |
+
"eval_samples_per_second": 194.222,
|
| 546 |
+
"eval_steps_per_second": 3.108,
|
| 547 |
+
"eval_within_var": 1.088273286819458,
|
| 548 |
+
"eval_wth_loss": 4.5452433568016204e-05,
|
| 549 |
+
"step": 7168
|
| 550 |
+
}
|
| 551 |
+
],
|
| 552 |
+
"logging_steps": 256,
|
| 553 |
+
"max_steps": 96210,
|
| 554 |
+
"num_input_tokens_seen": 0,
|
| 555 |
+
"num_train_epochs": 1,
|
| 556 |
+
"save_steps": 1024,
|
| 557 |
+
"stateful_callbacks": {
|
| 558 |
+
"TrainerControl": {
|
| 559 |
+
"args": {
|
| 560 |
+
"should_epoch_stop": false,
|
| 561 |
+
"should_evaluate": false,
|
| 562 |
+
"should_log": false,
|
| 563 |
+
"should_save": true,
|
| 564 |
+
"should_training_stop": false
|
| 565 |
+
},
|
| 566 |
+
"attributes": {}
|
| 567 |
+
}
|
| 568 |
+
},
|
| 569 |
+
"total_flos": 0.0,
|
| 570 |
+
"train_batch_size": 64,
|
| 571 |
+
"trial_name": null,
|
| 572 |
+
"trial_params": null
|
| 573 |
+
}
|
checkpoints-v3/checkpoint-7168/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9d4bda444d2217dfb8c9657e282f4547df7c92efeb7c6abde9e602028d4dde0a
|
| 3 |
+
size 5777
|