Training in progress, epoch 1

config.json

@@ -1,5 +1,5 @@
 {
-  "_name_or_path": "flaubert/flaubert_base_cased",
+  "_name_or_path": "flaubert/flaubert_small_cased",
   "amp": 1,
   "architectures": [
     "FlaubertForSequenceClassification"
@@ -12,7 +12,7 @@
   "causal": false,
   "clip_grad_norm": 5,
   "dropout": 0.1,
-  "emb_dim": 768,
+  "emb_dim": 512,
   "embed_init_std": 0.02209708691207961,
   "encoder_only": true,
   "end_n_top": 5,
@@ -48,8 +48,8 @@
   "langs": [
     "fr"
   ],
-  "layer_norm_eps": 1e-12,
-  "layerdrop": 0.0,
+  "layer_norm_eps": 1e-06,
+  "layerdrop": 0.2,
   "lg_sampling_factor": -1,
   "lgs": "fr",
   "mask_index": 5,
@@ -64,12 +64,12 @@
     ]
   ],
   "model_type": "flaubert",
-  "n_heads": 12,
+  "n_heads": 8,
   "n_langs": 1,
-  "n_layers": 12,
+  "n_layers": 6,
   "pad_index": 2,
   "pad_token_id": 2,
-  "pre_norm": false,
+  "pre_norm": true,
   "problem_type": "single_label_classification",
   "sample_alpha": 0,
   "share_inout_emb": true,

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0851904f9ef130b48a14ce3a0854a8aaeb731d368fc04930f22eb3256b2001c2
-size 553012965
+oid sha256:a4d40f5bc35e795c24e9f91d33a011ec9b51758f8aae890354a1c5df50757036
+size 217510949

runs/May16_16-37-02_473a1c8172e8/1684255259.427329/events.out.tfevents.1684255259.473a1c8172e8.4677.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f3c42ab47f4c9053355b192937c56ec2a06a7f87641f8c22d096b5443a3e63d2
+size 5898

runs/May16_16-37-02_473a1c8172e8/events.out.tfevents.1684255259.473a1c8172e8.4677.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0112f132f0b22cd172d497c85056335b0e8b769a3cdd0ef55ab6816d43a5742a
+size 5536

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7fffa31dafbd24e0ec9011f554292921a1e83e05e6129da37f6fde4e02ac83e3
+oid sha256:e7412c7dbbef8b2b0bdca3e0ca2bfc9b3b3a8a004c3c059ef73c23070be84513
 size 3899