Training in progress, step 12000, checkpoint

checkpoint-12000/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c4ae09d47b5295ce46ddd3b48cc736ad9e206efb7505dbecd7ba877c6f6db31c
+oid sha256:59ce8f38601fbd6eee49a901e382da7d78ebce507e4e265c62d6aafebca365e2
 size 51007160

checkpoint-12000/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:74eb6edf5a00642fc4a2785cff160a15179ba255a6a749f599217d8908e2b9d6
+oid sha256:03d99881f84e0354e580b4d33490efe5733785e6f803651a550a5c9e53c91454
 size 102078202

checkpoint-12000/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:56f44917d6c9223960f0046add7487286a53bb7d7d42fe253c3377466ea3e315
+oid sha256:e8987f95f1cceab22386ab9bede80f931023ad982dc5dee03b613fe9d1cd18e2
 size 14308

checkpoint-12000/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a300d405a31ed9a6a3aacb8615e61e340e96b3c687b201bcf6ed4ae9f299ec86
+oid sha256:5463b724c2cd6909cd79c85b289d00a369e4fb6bb042482b75620c59845c7360
 size 1000

checkpoint-12000/trainer_state.json

@@ -1,7 +1,7 @@
 {
-  "best_metric": 5.460005760192871,
-  "best_model_checkpoint": "/home/p318482/babyLM_controlled/models_trained/en_clm/wikipedia_30/checkpoint-12000",
-  "epoch": 6.310807257428346,
+  "best_metric": 4.517208576202393,
+  "best_model_checkpoint": "/home/p318482/babyLM_controlled/models_trained/fr_clm/wikipedia_30/checkpoint-12000",
+  "epoch": 12.882447665056361,
   "eval_steps": 2000,
   "global_step": 12000,
   "is_hyper_param_search": false,
@@ -9,79 +9,79 @@
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 1.0518012095713911,
-      "eval_loss": 7.518920421600342,
-      "eval_runtime": 2.2399,
-      "eval_samples_per_second": 1422.808,
-      "eval_steps_per_second": 89.288,
+      "epoch": 2.1470746108427265,
+      "eval_loss": 7.138043403625488,
+      "eval_runtime": 0.8415,
+      "eval_samples_per_second": 1273.93,
+      "eval_steps_per_second": 79.621,
       "step": 2000
     },
     {
-      "epoch": 2.1036024191427822,
-      "grad_norm": 1.2947496175765991,
+      "epoch": 4.294149221685453,
+      "grad_norm": 1.455039381980896,
       "learning_rate": 1e-05,
-      "loss": 7.6141,
+      "loss": 7.213,
       "step": 4000
     },
     {
-      "epoch": 2.1036024191427822,
-      "eval_loss": 6.568221092224121,
-      "eval_runtime": 2.1236,
-      "eval_samples_per_second": 1500.783,
-      "eval_steps_per_second": 94.182,
+      "epoch": 4.294149221685453,
+      "eval_loss": 5.8541717529296875,
+      "eval_runtime": 0.7753,
+      "eval_samples_per_second": 1382.676,
+      "eval_steps_per_second": 86.417,
       "step": 4000
     },
     {
-      "epoch": 3.155403628714173,
-      "eval_loss": 6.16925573348999,
-      "eval_runtime": 2.1624,
-      "eval_samples_per_second": 1473.825,
-      "eval_steps_per_second": 92.49,
+      "epoch": 6.4412238325281805,
+      "eval_loss": 5.403579235076904,
+      "eval_runtime": 0.7542,
+      "eval_samples_per_second": 1421.437,
+      "eval_steps_per_second": 88.84,
       "step": 6000
     },
     {
-      "epoch": 4.2072048382855645,
-      "grad_norm": 2.330080270767212,
-      "learning_rate": 2e-05,
-      "loss": 6.2396,
+      "epoch": 8.588298443370906,
+      "grad_norm": 3.117489814758301,
+      "learning_rate": 1.9997500000000003e-05,
+      "loss": 5.4304,
       "step": 8000
     },
     {
-      "epoch": 4.2072048382855645,
-      "eval_loss": 5.901575088500977,
-      "eval_runtime": 2.1326,
-      "eval_samples_per_second": 1494.388,
-      "eval_steps_per_second": 93.78,
+      "epoch": 8.588298443370906,
+      "eval_loss": 5.049880504608154,
+      "eval_runtime": 0.7597,
+      "eval_samples_per_second": 1411.127,
+      "eval_steps_per_second": 88.195,
       "step": 8000
     },
     {
-      "epoch": 5.259006047856955,
-      "eval_loss": 5.67369270324707,
-      "eval_runtime": 2.1364,
-      "eval_samples_per_second": 1491.788,
-      "eval_steps_per_second": 93.617,
+      "epoch": 10.735373054213634,
+      "eval_loss": 4.760603427886963,
+      "eval_runtime": 0.7624,
+      "eval_samples_per_second": 1406.053,
+      "eval_steps_per_second": 87.878,
       "step": 10000
     },
     {
-      "epoch": 6.310807257428346,
-      "grad_norm": 2.565220594406128,
-      "learning_rate": 2.99925e-05,
-      "loss": 5.7217,
+      "epoch": 12.882447665056361,
+      "grad_norm": 2.9511141777038574,
+      "learning_rate": 2.9995e-05,
+      "loss": 4.771,
       "step": 12000
     },
     {
-      "epoch": 6.310807257428346,
-      "eval_loss": 5.460005760192871,
-      "eval_runtime": 2.1835,
-      "eval_samples_per_second": 1459.563,
-      "eval_steps_per_second": 91.595,
+      "epoch": 12.882447665056361,
+      "eval_loss": 4.517208576202393,
+      "eval_runtime": 0.764,
+      "eval_samples_per_second": 1403.097,
+      "eval_steps_per_second": 87.694,
       "step": 12000
     }
   ],
   "logging_steps": 4000,
   "max_steps": 100000,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 53,
+  "num_train_epochs": 108,
   "save_steps": 4000,
   "stateful_callbacks": {
     "TrainerControl": {
@@ -95,7 +95,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 3102067886653440.0,
+  "total_flos": 3102019405479936.0,
   "train_batch_size": 16,
   "trial_name": null,
   "trial_params": null

checkpoint-12000/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ab68434ec7156d3a63e4783f824871a48d62c7c2fdcc831fdd5471c5f4aec7fe
+oid sha256:26102d0ac750ce531ebda9d7577ccea6c8ecbc93b04bdff23226e376dd5609a2
 size 5368