Training in progress, step 350, checkpoint

checkpoint-350/README.md

@@ -216,4 +216,4 @@ The following `bitsandbytes` quantization config was used during training:
 ### Framework versions
 
 
-- PEFT 0.6.0.dev0
+- PEFT 0.6.0

checkpoint-350/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:46093d76084cfdd856f08f94ed0dcd94496570686ef524936a11cf5a275d3cb8
+size 9859800

checkpoint-350/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3d20d85de5a909b6f669078080271b66fa2a018a9b314858a4f506a8a513e6cf
+oid sha256:3e627e285c7e08976e52c8884d24b96e3a27c7abc2309f7c136d04c1e4faf206
 size 42788

checkpoint-350/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0d4b8691751fe847b2c6d8c32cee34753910b8f8078ee0ab10328143ca8702d0
+oid sha256:aec31b53cb67f83ff597fa948b44d66eb900911995b5a752d113820f9eae0a44
 size 14244

checkpoint-350/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:eec15d083a9e53d5b5a92a0d271f0e7ed80eb2a6f5e825327a3c27ef830bf4aa
+oid sha256:df9d1b31b2857865711ef299f3ee7fef317f5b1be2f0b867dadb5c691f3187fd
 size 1064

checkpoint-350/special_tokens_map.json

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

checkpoint-350/tokenizer_config.json

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

checkpoint-350/trainer_state.json

@@ -1,7 +1,7 @@
 {
-  "best_metric": 0.5725014805793762,
-  "best_model_checkpoint": "output/checkpoint-100",
-  "epoch": 6.60377358490566,
+  "best_metric": 0.5582989454269409,
+  "best_model_checkpoint": "output/checkpoint-200",
+  "epoch": 0.2807580467261606,
   "eval_steps": 50,
   "global_step": 350,
   "is_hyper_param_search": false,
@@ -9,122 +9,122 @@
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.02,
-      "learning_rate": 1.234567901234568e-06,
-      "loss": 0.4437,
+      "epoch": 0.0,
+      "learning_rate": 4.444444444444445e-07,
+      "loss": 1.0919,
       "step": 1
     },
     {
-      "epoch": 0.94,
-      "learning_rate": 5.679012345679012e-05,
-      "loss": 0.455,
+      "epoch": 0.04,
+      "learning_rate": 2e-05,
+      "loss": 0.7661,
       "step": 50
     },
     {
-      "epoch": 0.94,
-      "eval_accuracy": 0.68,
-      "eval_loss": 0.6171127557754517,
-      "eval_runtime": 29.0494,
-      "eval_samples_per_second": 6.024,
-      "eval_steps_per_second": 1.515,
+      "epoch": 0.04,
+      "eval_accuracy": 0.7142857142857143,
+      "eval_loss": 0.6954500079154968,
+      "eval_runtime": 28.4403,
+      "eval_samples_per_second": 6.153,
+      "eval_steps_per_second": 1.547,
       "step": 50
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 9.59349593495935e-05,
-      "loss": 0.4555,
+      "epoch": 0.08,
+      "learning_rate": 4.222222222222222e-05,
+      "loss": 0.6119,
       "step": 100
     },
     {
-      "epoch": 1.89,
-      "eval_accuracy": 0.7257142857142858,
-      "eval_loss": 0.5725014805793762,
-      "eval_runtime": 28.7986,
-      "eval_samples_per_second": 6.077,
-      "eval_steps_per_second": 1.528,
+      "epoch": 0.08,
+      "eval_accuracy": 0.7314285714285714,
+      "eval_loss": 0.5819395780563354,
+      "eval_runtime": 28.3595,
+      "eval_samples_per_second": 6.171,
+      "eval_steps_per_second": 1.552,
       "step": 100
     },
     {
-      "epoch": 2.83,
-      "learning_rate": 8.265582655826559e-05,
-      "loss": 0.455,
+      "epoch": 0.12,
+      "learning_rate": 6.444444444444446e-05,
+      "loss": 0.5869,
       "step": 150
     },
     {
-      "epoch": 2.83,
+      "epoch": 0.12,
       "eval_accuracy": 0.7371428571428571,
-      "eval_loss": 0.5729417204856873,
-      "eval_runtime": 28.9054,
-      "eval_samples_per_second": 6.054,
-      "eval_steps_per_second": 1.522,
+      "eval_loss": 0.6004045605659485,
+      "eval_runtime": 28.3043,
+      "eval_samples_per_second": 6.183,
+      "eval_steps_per_second": 1.555,
       "step": 150
     },
     {
-      "epoch": 3.77,
-      "learning_rate": 6.910569105691057e-05,
-      "loss": 0.4718,
+      "epoch": 0.16,
+      "learning_rate": 8.666666666666667e-05,
+      "loss": 0.588,
       "step": 200
     },
     {
-      "epoch": 3.77,
-      "eval_accuracy": 0.6685714285714286,
-      "eval_loss": 0.6322054862976074,
-      "eval_runtime": 28.8878,
-      "eval_samples_per_second": 6.058,
-      "eval_steps_per_second": 1.523,
+      "epoch": 0.16,
+      "eval_accuracy": 0.76,
+      "eval_loss": 0.5582989454269409,
+      "eval_runtime": 28.3002,
+      "eval_samples_per_second": 6.184,
+      "eval_steps_per_second": 1.555,
       "step": 200
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 5.555555555555556e-05,
-      "loss": 0.4348,
+      "epoch": 0.2,
+      "learning_rate": 9.80411361410382e-05,
+      "loss": 0.5701,
       "step": 250
     },
     {
-      "epoch": 4.72,
-      "eval_accuracy": 0.7257142857142858,
-      "eval_loss": 0.5733612179756165,
-      "eval_runtime": 28.5713,
-      "eval_samples_per_second": 6.125,
-      "eval_steps_per_second": 1.54,
+      "epoch": 0.2,
+      "eval_accuracy": 0.7142857142857143,
+      "eval_loss": 0.6456161141395569,
+      "eval_runtime": 28.2527,
+      "eval_samples_per_second": 6.194,
+      "eval_steps_per_second": 1.557,
       "step": 250
     },
     {
-      "epoch": 5.66,
-      "learning_rate": 4.2005420054200545e-05,
-      "loss": 0.4353,
+      "epoch": 0.24,
+      "learning_rate": 9.31439764936337e-05,
+      "loss": 0.5652,
       "step": 300
     },
     {
-      "epoch": 5.66,
-      "eval_accuracy": 0.6628571428571428,
-      "eval_loss": 0.6271094679832458,
-      "eval_runtime": 26.5252,
-      "eval_samples_per_second": 6.597,
-      "eval_steps_per_second": 1.659,
+      "epoch": 0.24,
+      "eval_accuracy": 0.72,
+      "eval_loss": 0.6802676916122437,
+      "eval_runtime": 28.3389,
+      "eval_samples_per_second": 6.175,
+      "eval_steps_per_second": 1.553,
       "step": 300
     },
     {
-      "epoch": 6.6,
-      "learning_rate": 2.8455284552845528e-05,
-      "loss": 0.4261,
+      "epoch": 0.28,
+      "learning_rate": 8.82468168462292e-05,
+      "loss": 0.5936,
       "step": 350
     },
     {
-      "epoch": 6.6,
-      "eval_accuracy": 0.6685714285714286,
-      "eval_loss": 0.6181718111038208,
-      "eval_runtime": 26.5098,
-      "eval_samples_per_second": 6.601,
-      "eval_steps_per_second": 1.66,
+      "epoch": 0.28,
+      "eval_accuracy": 0.7542857142857143,
+      "eval_loss": 0.5668751001358032,
+      "eval_runtime": 28.2611,
+      "eval_samples_per_second": 6.192,
+      "eval_steps_per_second": 1.557,
       "step": 350
     }
   ],
   "logging_steps": 50,
-  "max_steps": 450,
-  "num_train_epochs": 9,
+  "max_steps": 1246,
+  "num_train_epochs": 1,
   "save_steps": 50,
-  "total_flos": 1.339927328299008e+17,
+  "total_flos": 1.2952272643915776e+17,
   "trial_name": null,
   "trial_params": null
 }

checkpoint-350/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d91c1e3ec7fc28d5ed2a9d783b7ded0a7affbbd029319e299d9244ac0f3fbc29
-size 4472
+oid sha256:b25ecbcf37659c2f593e1603dab54c3eaea43862c7f192a7e6df4c4dcc86d20d
+size 4600