Training in progress, step 2500

config.json

@@ -0,0 +1,44 @@
+{
+  "activation_dropout": 0.0,
+  "activation_function": "gelu",
+  "architectures": [
+    "CopyMechanismBart"
+  ],
+  "attention_dropout": 0.1,
+  "bos_token_id": 0,
+  "classifier_dropout": 0.0,
+  "d_model": 512,
+  "decoder_attention_heads": 8,
+  "decoder_ffn_dim": 2048,
+  "decoder_layerdrop": 0.0,
+  "decoder_layers": 6,
+  "decoder_start_token_id": 2,
+  "dropout": 0.1,
+  "encoder_attention_heads": 8,
+  "encoder_ffn_dim": 2048,
+  "encoder_layerdrop": 0.0,
+  "encoder_layers": 6,
+  "eos_token_id": 2,
+  "forced_eos_token_id": 2,
+  "id2label": {
+    "0": "LABEL_0",
+    "1": "LABEL_1",
+    "2": "LABEL_2"
+  },
+  "init_std": 0.02,
+  "is_encoder_decoder": true,
+  "label2id": {
+    "LABEL_0": 0,
+    "LABEL_1": 1,
+    "LABEL_2": 2
+  },
+  "max_position_embeddings": 512,
+  "model_type": "bart",
+  "num_hidden_layers": 6,
+  "pad_token_id": 1,
+  "scale_embedding": false,
+  "torch_dtype": "float32",
+  "transformers_version": "4.50.3",
+  "use_cache": true,
+  "vocab_size": 103
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3861c43adc539862ca848faf6f537a910953f6a55942de1b12b6bf3d81841542
+size 178908572

runs/Apr08_06-56-07_051cd242831b/events.out.tfevents.1744095379.051cd242831b.522.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:416176c474f5d4c425ab5b49dde6c9cebfdba3ebed80708d859dc8d213500a18
+size 5326

runs/Apr08_07-01-34_051cd242831b/events.out.tfevents.1744095695.051cd242831b.4215.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3cc884b2dd727770c9af624348a753b3e184d89c45c158b18ce6e9e7fa0f27b3
+size 5808

special_tokens.json

@@ -0,0 +1 @@
+["<pad>", "<s>", "</s>", "<unk>"]

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fa40c0dfc4d9619c96d0850ca6331f2d16d32941e6c6c586dc1464e4c273e9d6
+size 5304

vocab.json

@@ -0,0 +1 @@
+{"<pad>": 0, "<s>": 1, "</s>": 2, "<unk>": 3, "\u0905": 4, "\u0906": 5, "\u0907": 6, "\u0908": 7, "\u0909": 8, "\u090a": 9, "\u090b": 10, "\u0960": 11, "\u090c": 12, "\u0961": 13, "\u090e": 14, "\u090f": 15, "\u0910": 16, "\u0913": 17, "\u0914": 18, "\u0915": 19, "\u0916": 20, "\u0917": 21, "\u0918": 22, "\u0919": 23, "\u091a": 24, "\u091b": 25, "\u091c": 26, "\u091d": 27, "\u091e": 28, "\u091f": 29, "\u0920": 30, "\u0921": 93, "\u0922": 101, "\u0923": 33, "\u0924": 34, "\u0925": 35, "\u0926": 36, "\u0927": 37, "\u0928": 38, "\u092a": 39, "\u092b": 99, "\u092c": 41, "\u092d": 42, "\u092e": 43, "\u092f": 44, "\u0930": 45, "\u0932": 46, "\u0935": 47, "\u0936": 48, "\u0937": 49, "\u0938": 50, "\u0939": 51, "\u094d": 52, "\u093e": 53, "\u093f": 54, "\u0940": 55, "\u0941": 56, "\u0942": 57, "\u0943": 58, "\u0944": 59, "\u0947": 60, "\u0948": 61, "\u094b": 62, "\u094c": 63, "\u0902": 86, "\u0903": 87, "\u0901": 67, "\u0964": 68, "\u0966": 69, "\u0967": 70, "\u0968": 71, "\u0969": 72, "\u096a": 73, "\u096b": 74, "\u096c": 75, "\u096d": 76, "\u096e": 77, "\u096f": 78, ",": 79, ".": 80, ";": 81, "?": 82, "!": 83, "\u093c": 102, "\u0945": 89, "\u094a": 88, "\u0912": 90, "\u093d": 91, "\u0949": 92, "\u0950": 95, "\u0965": 96, "\u0911": 97, "\u0931": 98}