Training in progress, step 500

.gitignore

@@ -0,0 +1 @@
+checkpoint-*/

config.json

@@ -0,0 +1,32 @@
+{
+  "architectures": [
+    "BigBirdForCausalLM"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "attention_type": "block_sparse",
+  "block_size": 64,
+  "bos_token_id": 1,
+  "classifier_dropout": null,
+  "eos_token_id": 2,
+  "hidden_act": "gelu_new",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 768,
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "is_decoder": true,
+  "layer_norm_eps": 1e-12,
+  "max_position_embeddings": 4096,
+  "model_type": "big_bird",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 12,
+  "num_random_blocks": 3,
+  "pad_token_id": 0,
+  "rescale_embeddings": false,
+  "sep_token_id": 66,
+  "torch_dtype": "float32",
+  "transformers_version": "4.24.0",
+  "type_vocab_size": 2,
+  "use_bias": true,
+  "use_cache": true,
+  "vocab_size": 40000
+}

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f55f85b43d485e9131b499b44f16582f7e1f873f30a3e5298987e1d0601a26d7
+size 480686657

runs/Jan31_10-58-33_tardis/1675159121.616173/events.out.tfevents.1675159121.tardis.1781751.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0ddad2bb9f1195ba788237bf45913fc43dd3aeb08829ed2bc900c420df6d9191
+size 5479

runs/Jan31_10-58-33_tardis/events.out.tfevents.1675159121.tardis.1781751.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6747c4c3f05e70ea4d462e9eaa7bc88565b36fb3806c8c69c982c6446226c078
+size 3759

runs/Jan31_10-59-47_tardis/1675159192.0490732/events.out.tfevents.1675159192.tardis.1782105.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:62e03eb4080ca453664ba313e23b231b65ed6e06614d65299e791d87a3e8f7e3
+size 5479

runs/Jan31_10-59-47_tardis/events.out.tfevents.1675159192.tardis.1782105.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9ebcc8751955e90278b9064fdf0aacc0cfb37fa4104bddf44071c03805e2d7a4
+size 3781

runs/Jan31_11-01-39_tardis/1675159304.356033/events.out.tfevents.1675159304.tardis.1783035.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f26f4891cd319865dc1f425d99510d6d3426a27d9a891317c4cad912512fd58f
+size 5479

runs/Jan31_11-01-39_tardis/events.out.tfevents.1675159304.tardis.1783035.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:27b50f37895469bcd0290fdf96bcb7162e5825d12ca6d2a14f07d0daf4d1303b
+size 3781

runs/Jan31_11-02-13_tardis/1675159337.6714537/events.out.tfevents.1675159337.tardis.1783627.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0e79ae0c89969c1bacdff4b512cdd2f4dafafded3155e0d0ac63912b735205bd
+size 5479

runs/Jan31_11-02-13_tardis/events.out.tfevents.1675159337.tardis.1783627.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5428bbe33712a06f870a918617d40a994563e77581c6d19e750373140329c0b6
+size 3938

special_tokens_map.json

@@ -0,0 +1,9 @@
+{
+  "bos_token": "<s>",
+  "cls_token": "<cls>",
+  "eos_token": "</s>",
+  "mask_token": "<mask>",
+  "pad_token": "</s>",
+  "sep_token": "<sep>",
+  "unk_token": "<unk>"
+}

tokenizer_config.json

@@ -0,0 +1,5 @@
+{
+  "name_or_path": "/home/ma/s/schroederl/XNEXT/xnext/data/tokenizer_fast",
+  "special_tokens_map_file": "/home/ma/s/schroederl/XNEXT/xnext/data/tokenizer_fast/special_tokens_map.json",
+  "tokenizer_class": "PreTrainedTokenizerFast"
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:295fdccb4b1230676641397539dc00887e6b1de30611c21910bcf99fcd9cc6d6
+size 3451