Training in progress, step 4000, checkpoint

checkpoint-4000/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b1a41c0c9c471f5fe4a8545c6c432371aed3a1aab6a5bbc6b4fb1e2ab6abe684
+oid sha256:8a719ed1c47d7040d8c271fbddafe3817c627ea3f7472899b0a886c1b1a5af71
 size 51007160

checkpoint-4000/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:018a342ed609cfddbf13ea2a74f8e4c03e1583172daff1c1913e53bfc016f4d5
+oid sha256:c904ad9f6714d7b21b5751d0bb1f8c15715193ac13551d609002c5ee2d2a046a
 size 102078202

checkpoint-4000/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f0ec3cad708a6639c46b6e1d7ec67907f5136cf0dd75a03fc8d9c1548124fc18
+oid sha256:873e793df4e0f9482a3f11eebaec32f07e9f61801291da1c3626b458bb5711bf
 size 14244

checkpoint-4000/trainer_state.json

@@ -1,7 +1,7 @@
 {
-  "best_metric": 5.101934432983398,
-  "best_model_checkpoint": "/home/p318482/babyLM_controlled/models_trained/fr_clm/childes_42/checkpoint-4000",
-  "epoch": 6.289308176100629,
+  "best_metric": 5.874638557434082,
+  "best_model_checkpoint": "/home/p318482/babyLM_controlled/models_trained/de_clm/childes_42/checkpoint-4000",
+  "epoch": 4.1928721174004195,
   "eval_steps": 2000,
   "global_step": 4000,
   "is_hyper_param_search": false,
@@ -9,33 +9,33 @@
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 3.1446540880503147,
-      "eval_loss": 6.666613578796387,
-      "eval_runtime": 0.6599,
-      "eval_samples_per_second": 1380.413,
-      "eval_steps_per_second": 86.37,
+      "epoch": 2.0964360587002098,
+      "eval_loss": 7.0908098220825195,
+      "eval_runtime": 0.9834,
+      "eval_samples_per_second": 1398.23,
+      "eval_steps_per_second": 87.453,
       "step": 2000
     },
     {
-      "epoch": 6.289308176100629,
-      "grad_norm": 1.431768536567688,
+      "epoch": 4.1928721174004195,
+      "grad_norm": 1.4699604511260986,
       "learning_rate": 1e-05,
-      "loss": 6.5315,
+      "loss": 6.9765,
       "step": 4000
     },
     {
-      "epoch": 6.289308176100629,
-      "eval_loss": 5.101934432983398,
-      "eval_runtime": 0.6406,
-      "eval_samples_per_second": 1422.145,
-      "eval_steps_per_second": 88.982,
+      "epoch": 4.1928721174004195,
+      "eval_loss": 5.874638557434082,
+      "eval_runtime": 0.973,
+      "eval_samples_per_second": 1413.214,
+      "eval_steps_per_second": 88.39,
       "step": 4000
     }
   ],
   "logging_steps": 4000,
   "max_steps": 100000,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 158,
+  "num_train_epochs": 105,
   "save_steps": 4000,
   "stateful_callbacks": {
     "TrainerControl": {
@@ -49,7 +49,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 1034168072404992.0,
+  "total_flos": 1034264024727552.0,
   "train_batch_size": 16,
   "trial_name": null,
   "trial_params": null

checkpoint-4000/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e6d9a18dff8e8501ed8e1286b71e34a5ee0478c42b015fd65b8981e86e58a8c2
+oid sha256:7e7557b0dc8ef07c9db255e1167d98819525b043deb4934f580e21d716914111
 size 5368