Training in progress, epoch 2

config.json

@@ -1,5 +1,5 @@
 {
-  "_name_or_path": "facebook/m2m100_418M",
+  "_name_or_path": "facebook/m2m100_1.2B",
   "activation_dropout": 0.0,
   "activation_function": "relu",
   "architectures": [
@@ -9,16 +9,16 @@
   "bos_token_id": 0,
   "d_model": 1024,
   "decoder_attention_heads": 16,
-  "decoder_ffn_dim": 4096,
+  "decoder_ffn_dim": 8192,
   "decoder_layerdrop": 0.05,
-  "decoder_layers": 12,
+  "decoder_layers": 24,
   "decoder_start_token_id": 2,
-  "dropout": 0.1,
+  "dropout": 0.2,
   "early_stopping": null,
   "encoder_attention_heads": 16,
-  "encoder_ffn_dim": 4096,
+  "encoder_ffn_dim": 8192,
   "encoder_layerdrop": 0.05,
-  "encoder_layers": 12,
+  "encoder_layers": 24,
   "eos_token_id": 2,
   "gradient_checkpointing": false,
   "init_std": 0.02,
@@ -27,7 +27,7 @@
   "max_position_embeddings": 1024,
   "model_type": "m2m_100",
   "num_beams": null,
-  "num_hidden_layers": 12,
+  "num_hidden_layers": 24,
   "pad_token_id": 1,
   "scale_embedding": true,
   "torch_dtype": "float32",

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:85acc9eb703100d1a2ee205ce1ea02102af093a011deb454ae8131ae787e0135
-size 1935681888
+oid sha256:6531e6a81a0c7e8cf16e1545a864e6a3fc7f143f7360bad61ad95fe3290dd95e
+size 4958000808

tokenizer_config.json

@@ -940,7 +940,7 @@
   "eos_token": "</s>",
   "extra_special_tokens": {},
   "language_codes": "m2m100",
-  "model_max_length": 1024,
+  "model_max_length": 1000000000000000019884624838656,
   "num_madeup_words": 8,
   "pad_token": "<pad>",
   "sep_token": "</s>",

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:da8f1f3633486f36fa576c1959d1b6853e7f9034c522dbfeae1d94ebe1ae8206
+oid sha256:714c92beaf552f27f67a23b03b950a9a013d364e90f6924abd63211e7cb5fe8e
 size 5496