Training in progress, step 400, checkpoint

checkpoint-400/adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:859d28acbf3f93a78c35051d24de0523704e40839fa3d04fc2bba30158720969
+oid sha256:78a89d73e9d5b54bf63e75b052e14bf0ff3fa43e3b2c5521b2965d8fd0ef2913
 size 19690328

checkpoint-400/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5e01a945972f87560918e64274ea672d97ae0ed4d4d15c673798905f3ec52b7c
+oid sha256:da8a228dc9fc24d7f01850b9b5fa58e2329ee4efeccf99baa37ba367bf97ea09
 size 38087162

checkpoint-400/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e9e1427ce5cf3cbbccc6910f9762f2d8226e590c46462a05004c7de49f683a1f
+oid sha256:52db077dd0a6b5083f8d03a5879b31c139f9eb41dca2d8b4dc332a6c18ec6316
 size 14244

checkpoint-400/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0c53722d710c782032164ef552743070347f4abb50a523a268cff04be6ecd4e4
+oid sha256:9230dc0493c33f4081093f9d8f0d7dfe87618eae9f6d0d25007c90dfedb7c988
 size 1064

checkpoint-400/special_tokens_map.json

@@ -2,6 +2,6 @@
   "bos_token": "<s>",
   "eos_token": "</s>",
   "pad_token": "<pad>",
-  "sep_token": "\n\n",
+  "sep_token": "<s>",
   "unk_token": "<unk>"
 }

checkpoint-400/tokenizer_config.json

@@ -39,7 +39,7 @@
   "eos_token": "</s>",
   "model_max_length": 1000000000000000019884624838656,
   "pad_token": "<pad>",
-  "sep_token": "\n\n",
+  "sep_token": "<s>",
   "tokenizer_class": "BloomTokenizer",
   "unk_token": "<unk>"
 }

checkpoint-400/trainer_state.json

@@ -1,7 +1,7 @@
 {
-  "best_metric": 0.6207379102706909,
-  "best_model_checkpoint": "output/checkpoint-100",
-  "epoch": 1.6692749087115284,
+  "best_metric": 0.47385331988334656,
+  "best_model_checkpoint": "output/checkpoint-150",
+  "epoch": 0.32844093195114443,
   "eval_steps": 50,
   "global_step": 400,
   "is_hyper_param_search": false,
@@ -11,135 +11,135 @@
     {
       "epoch": 0.0,
       "learning_rate": 0.0,
-      "loss": 0.4694,
+      "loss": 0.8903,
       "step": 1
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 1.25e-05,
-      "loss": 0.3006,
+      "epoch": 0.04,
+      "learning_rate": 1.2222222222222224e-05,
+      "loss": 0.7189,
       "step": 50
     },
     {
-      "epoch": 0.21,
-      "eval_accuracy": 0.6971428571428572,
-      "eval_loss": 0.6772400140762329,
-      "eval_runtime": 28.6241,
-      "eval_samples_per_second": 6.114,
-      "eval_steps_per_second": 1.537,
+      "epoch": 0.04,
+      "eval_accuracy": 0.8228571428571428,
+      "eval_loss": 0.5415178537368774,
+      "eval_runtime": 28.5756,
+      "eval_samples_per_second": 6.124,
+      "eval_steps_per_second": 1.54,
       "step": 50
     },
     {
-      "epoch": 0.42,
-      "learning_rate": 1.8597560975609757e-05,
-      "loss": 0.1715,
+      "epoch": 0.08,
+      "learning_rate": 1.8658536585365855e-05,
+      "loss": 0.5723,
       "step": 100
     },
     {
-      "epoch": 0.42,
-      "eval_accuracy": 0.8228571428571428,
-      "eval_loss": 0.6207379102706909,
-      "eval_runtime": 28.9688,
-      "eval_samples_per_second": 6.041,
-      "eval_steps_per_second": 1.519,
+      "epoch": 0.08,
+      "eval_accuracy": 0.8,
+      "eval_loss": 0.49520450830459595,
+      "eval_runtime": 28.516,
+      "eval_samples_per_second": 6.137,
+      "eval_steps_per_second": 1.543,
       "step": 100
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 1.554878048780488e-05,
-      "loss": 0.1573,
+      "epoch": 0.12,
+      "learning_rate": 1.5609756097560978e-05,
+      "loss": 0.5122,
       "step": 150
     },
     {
-      "epoch": 0.63,
-      "eval_accuracy": 0.76,
-      "eval_loss": 0.7828966379165649,
-      "eval_runtime": 26.1173,
-      "eval_samples_per_second": 6.701,
-      "eval_steps_per_second": 1.685,
+      "epoch": 0.12,
+      "eval_accuracy": 0.7885714285714286,
+      "eval_loss": 0.47385331988334656,
+      "eval_runtime": 28.2004,
+      "eval_samples_per_second": 6.206,
+      "eval_steps_per_second": 1.56,
       "step": 150
     },
     {
-      "epoch": 0.83,
-      "learning_rate": 1.25e-05,
-      "loss": 0.1264,
+      "epoch": 0.16,
+      "learning_rate": 1.2560975609756098e-05,
+      "loss": 0.4831,
       "step": 200
     },
     {
-      "epoch": 0.83,
-      "eval_accuracy": 0.8171428571428572,
-      "eval_loss": 0.7038331627845764,
-      "eval_runtime": 28.7177,
-      "eval_samples_per_second": 6.094,
-      "eval_steps_per_second": 1.532,
+      "epoch": 0.16,
+      "eval_accuracy": 0.7942857142857143,
+      "eval_loss": 0.49147841334342957,
+      "eval_runtime": 28.6363,
+      "eval_samples_per_second": 6.111,
+      "eval_steps_per_second": 1.537,
       "step": 200
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 9.451219512195122e-06,
-      "loss": 0.1099,
+      "epoch": 0.21,
+      "learning_rate": 9.51219512195122e-06,
+      "loss": 0.5009,
       "step": 250
     },
     {
-      "epoch": 1.04,
-      "eval_accuracy": 0.72,
-      "eval_loss": 1.0845664739608765,
-      "eval_runtime": 26.1061,
-      "eval_samples_per_second": 6.703,
-      "eval_steps_per_second": 1.685,
+      "epoch": 0.21,
+      "eval_accuracy": 0.8,
+      "eval_loss": 0.501244068145752,
+      "eval_runtime": 27.6943,
+      "eval_samples_per_second": 6.319,
+      "eval_steps_per_second": 1.589,
       "step": 250
     },
     {
-      "epoch": 1.25,
-      "learning_rate": 6.402439024390244e-06,
-      "loss": 0.108,
+      "epoch": 0.25,
+      "learning_rate": 6.463414634146342e-06,
+      "loss": 0.4883,
       "step": 300
     },
     {
-      "epoch": 1.25,
-      "eval_accuracy": 0.7428571428571429,
-      "eval_loss": 1.1050784587860107,
-      "eval_runtime": 28.843,
-      "eval_samples_per_second": 6.067,
-      "eval_steps_per_second": 1.526,
+      "epoch": 0.25,
+      "eval_accuracy": 0.7885714285714286,
+      "eval_loss": 0.49123460054397583,
+      "eval_runtime": 28.0924,
+      "eval_samples_per_second": 6.229,
+      "eval_steps_per_second": 1.566,
       "step": 300
     },
     {
-      "epoch": 1.46,
-      "learning_rate": 3.3536585365853664e-06,
-      "loss": 0.0842,
+      "epoch": 0.29,
+      "learning_rate": 3.414634146341464e-06,
+      "loss": 0.4979,
       "step": 350
     },
     {
-      "epoch": 1.46,
-      "eval_accuracy": 0.7771428571428571,
-      "eval_loss": 0.9869961738586426,
-      "eval_runtime": 26.1189,
-      "eval_samples_per_second": 6.7,
-      "eval_steps_per_second": 1.685,
+      "epoch": 0.29,
+      "eval_accuracy": 0.8,
+      "eval_loss": 0.4914019703865051,
+      "eval_runtime": 28.7942,
+      "eval_samples_per_second": 6.078,
+      "eval_steps_per_second": 1.528,
       "step": 350
     },
     {
-      "epoch": 1.67,
-      "learning_rate": 3.0487804878048784e-07,
-      "loss": 0.0852,
+      "epoch": 0.33,
+      "learning_rate": 3.6585365853658536e-07,
+      "loss": 0.5208,
       "step": 400
     },
     {
-      "epoch": 1.67,
-      "eval_accuracy": 0.7885714285714286,
-      "eval_loss": 0.9283963441848755,
-      "eval_runtime": 26.127,
-      "eval_samples_per_second": 6.698,
-      "eval_steps_per_second": 1.684,
+      "epoch": 0.33,
+      "eval_accuracy": 0.8057142857142857,
+      "eval_loss": 0.49244898557662964,
+      "eval_runtime": 28.8556,
+      "eval_samples_per_second": 6.065,
+      "eval_steps_per_second": 1.525,
       "step": 400
     }
   ],
   "logging_steps": 50,
   "max_steps": 400,
-  "num_train_epochs": 2,
+  "num_train_epochs": 1,
   "save_steps": 50,
-  "total_flos": 1.4830236693344256e+17,
+  "total_flos": 1.4827795775901696e+17,
   "trial_name": null,
   "trial_params": null
 }

checkpoint-400/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5027c9d6c3122f21d433b1e532b4aebee600801d0d4723deb6d4671b22b3d2da
+oid sha256:090099aa07c672025c60437897a7a19f63d8d7bd4284af2562fc632695431956
 size 4600