add fidmrg

.gitattributes

@@ -33,3 +33,20 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+*/*/ptorch_model.bin filter=lfs diff=lfs merge=lfs -text
+fidmrg_random.bm25.fidcqg-w/checkpoint-10000/pytorch_model.bin filter=lfs diff=lfs merge=lfs -text
+fidmrg_random.bm25.fidcqg-w/checkpoint-5000/pytorch_model.bin filter=lfs diff=lfs merge=lfs -text
+fidmrg_random.bm25.fidcqg/checkpoint-10000/pytorch_model.bin filter=lfs diff=lfs merge=lfs -text
+fidmrg_random.bm25.fidcqg/checkpoint-5000/pytorch_model.bin filter=lfs diff=lfs merge=lfs -text
+fidmrg_answer.bm25.fidcqg-w/checkpoint-10000/pytorch_model.bin filter=lfs diff=lfs merge=lfs -text
+fidmrg_answer.bm25.fidcqg-w/checkpoint-5000/pytorch_model.bin filter=lfs diff=lfs merge=lfs -text
+fidmrg_enumerated.bm25.fidcqg-w/checkpoint-10000/pytorch_model.bin filter=lfs diff=lfs merge=lfs -text
+fidmrg_enumerated.bm25.fidcqg-w/checkpoint-5000/pytorch_model.bin filter=lfs diff=lfs merge=lfs -text
+fidmrg_random.bm25.fidcqg/checkpoint-10000/optimizer.pt filter=lfs diff=lfs merge=lfs -text
+fidmrg_random.bm25.fidcqg/checkpoint-5000/optimizer.pt filter=lfs diff=lfs merge=lfs -text
+fidmrg_answer.bm25.fidcqg-w/checkpoint-10000/optimizer.pt filter=lfs diff=lfs merge=lfs -text
+fidmrg_answer.bm25.fidcqg-w/checkpoint-5000/optimizer.pt filter=lfs diff=lfs merge=lfs -text
+fidmrg_enumerated.bm25.fidcqg-w/checkpoint-10000/optimizer.pt filter=lfs diff=lfs merge=lfs -text
+fidmrg_enumerated.bm25.fidcqg-w/checkpoint-5000/optimizer.pt filter=lfs diff=lfs merge=lfs -text
+fidmrg_random.bm25.fidcqg-w/checkpoint-10000/optimizer.pt filter=lfs diff=lfs merge=lfs -text
+fidmrg_random.bm25.fidcqg-w/checkpoint-5000/optimizer.pt filter=lfs diff=lfs merge=lfs -text

fidmrg_answer.bm25.fidcqg-w/checkpoint-10000/config.json

@@ -0,0 +1,62 @@
+{
+  "_name_or_path": "google/flan-t5-base",
+  "architectures": [
+    "FiDT5"
+  ],
+  "classifier_dropout": 0.0,
+  "d_ff": 2048,
+  "d_kv": 64,
+  "d_model": 768,
+  "decoder_start_token_id": 0,
+  "dense_act_fn": "gelu_new",
+  "dropout_rate": 0.1,
+  "eos_token_id": 1,
+  "feed_forward_proj": "gated-gelu",
+  "initializer_factor": 1.0,
+  "is_encoder_decoder": true,
+  "is_gated_act": true,
+  "layer_norm_epsilon": 1e-06,
+  "model_type": "t5",
+  "n_positions": 512,
+  "num_decoder_layers": 12,
+  "num_heads": 12,
+  "num_layers": 12,
+  "output_past": true,
+  "pad_token_id": 0,
+  "relative_attention_max_distance": 128,
+  "relative_attention_num_buckets": 32,
+  "task_specific_params": {
+    "summarization": {
+      "early_stopping": true,
+      "length_penalty": 2.0,
+      "max_length": 200,
+      "min_length": 30,
+      "no_repeat_ngram_size": 3,
+      "num_beams": 4,
+      "prefix": "summarize: "
+    },
+    "translation_en_to_de": {
+      "early_stopping": true,
+      "max_length": 300,
+      "num_beams": 4,
+      "prefix": "translate English to German: "
+    },
+    "translation_en_to_fr": {
+      "early_stopping": true,
+      "max_length": 300,
+      "num_beams": 4,
+      "prefix": "translate English to French: "
+    },
+    "translation_en_to_ro": {
+      "early_stopping": true,
+      "max_length": 300,
+      "num_beams": 4,
+      "prefix": "translate English to Romanian: "
+    }
+  },
+  "tie_word_embeddings": false,
+  "torch_dtype": "float32",
+  "transformers_version": "4.33.1",
+  "use_cache": true,
+  "vocab_size": 32128
+}

fidmrg_answer.bm25.fidcqg-w/checkpoint-10000/generation_config.json

@@ -0,0 +1,7 @@
+{
+  "_from_model_config": true,
+  "decoder_start_token_id": 0,
+  "eos_token_id": 1,
+  "pad_token_id": 0,
+  "transformers_version": "4.33.1"
+}

fidmrg_answer.bm25.fidcqg-w/checkpoint-10000/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6a9289bad16b8a763d35664e88bca3a333d8548754e4554fe732c63c7079b36b
+size 1980864314

fidmrg_answer.bm25.fidcqg-w/checkpoint-10000/pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:087f380635736f9eb5b8fbd098aa7157aa88ac3bc5099b1ffaa6993969b11d81
+size 990411190

fidmrg_answer.bm25.fidcqg-w/checkpoint-10000/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:01bffb6fd7cce7b2e1cf3044180ca98a74bce41013a965361396b448a436701a
+size 14244

fidmrg_answer.bm25.fidcqg-w/checkpoint-10000/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:07356cb668b8efc440ddb2f7f81f7da3152060e81d0229922fa3b30df3313f82
+size 1064

fidmrg_answer.bm25.fidcqg-w/checkpoint-10000/special_tokens_map.json

@@ -0,0 +1,107 @@
+{
+  "additional_special_tokens": [
+    "<extra_id_0>",
+    "<extra_id_1>",
+    "<extra_id_2>",
+    "<extra_id_3>",
+    "<extra_id_4>",
+    "<extra_id_5>",
+    "<extra_id_6>",
+    "<extra_id_7>",
+    "<extra_id_8>",
+    "<extra_id_9>",
+    "<extra_id_10>",
+    "<extra_id_11>",
+    "<extra_id_12>",
+    "<extra_id_13>",
+    "<extra_id_14>",
+    "<extra_id_15>",
+    "<extra_id_16>",
+    "<extra_id_17>",
+    "<extra_id_18>",
+    "<extra_id_19>",
+    "<extra_id_20>",
+    "<extra_id_21>",
+    "<extra_id_22>",
+    "<extra_id_23>",
+    "<extra_id_24>",
+    "<extra_id_25>",
+    "<extra_id_26>",
+    "<extra_id_27>",
+    "<extra_id_28>",
+    "<extra_id_29>",
+    "<extra_id_30>",
+    "<extra_id_31>",
+    "<extra_id_32>",
+    "<extra_id_33>",
+    "<extra_id_34>",
+    "<extra_id_35>",
+    "<extra_id_36>",
+    "<extra_id_37>",
+    "<extra_id_38>",
+    "<extra_id_39>",
+    "<extra_id_40>",
+    "<extra_id_41>",
+    "<extra_id_42>",
+    "<extra_id_43>",
+    "<extra_id_44>",
+    "<extra_id_45>",
+    "<extra_id_46>",
+    "<extra_id_47>",
+    "<extra_id_48>",
+    "<extra_id_49>",
+    "<extra_id_50>",
+    "<extra_id_51>",
+    "<extra_id_52>",
+    "<extra_id_53>",
+    "<extra_id_54>",
+    "<extra_id_55>",
+    "<extra_id_56>",
+    "<extra_id_57>",
+    "<extra_id_58>",
+    "<extra_id_59>",
+    "<extra_id_60>",
+    "<extra_id_61>",
+    "<extra_id_62>",
+    "<extra_id_63>",
+    "<extra_id_64>",
+    "<extra_id_65>",
+    "<extra_id_66>",
+    "<extra_id_67>",
+    "<extra_id_68>",
+    "<extra_id_69>",
+    "<extra_id_70>",
+    "<extra_id_71>",
+    "<extra_id_72>",
+    "<extra_id_73>",
+    "<extra_id_74>",
+    "<extra_id_75>",
+    "<extra_id_76>",
+    "<extra_id_77>",
+    "<extra_id_78>",
+    "<extra_id_79>",
+    "<extra_id_80>",
+    "<extra_id_81>",
+    "<extra_id_82>",
+    "<extra_id_83>",
+    "<extra_id_84>",
+    "<extra_id_85>",
+    "<extra_id_86>",
+    "<extra_id_87>",
+    "<extra_id_88>",
+    "<extra_id_89>",
+    "<extra_id_90>",
+    "<extra_id_91>",
+    "<extra_id_92>",
+    "<extra_id_93>",
+    "<extra_id_94>",
+    "<extra_id_95>",
+    "<extra_id_96>",
+    "<extra_id_97>",
+    "<extra_id_98>",
+    "<extra_id_99>"
+  ],
+  "eos_token": "</s>",
+  "pad_token": "<pad>",
+  "unk_token": "<unk>"
+}

fidmrg_answer.bm25.fidcqg-w/checkpoint-10000/spiece.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d60acb128cf7b7f2536e8f38a5b18a05535c9e14c7a355904270e15b0945ea86
+size 791656

fidmrg_answer.bm25.fidcqg-w/checkpoint-10000/tokenizer_config.json

@@ -0,0 +1,112 @@
+{
+  "additional_special_tokens": [
+    "<extra_id_0>",
+    "<extra_id_1>",
+    "<extra_id_2>",
+    "<extra_id_3>",
+    "<extra_id_4>",
+    "<extra_id_5>",
+    "<extra_id_6>",
+    "<extra_id_7>",
+    "<extra_id_8>",
+    "<extra_id_9>",
+    "<extra_id_10>",
+    "<extra_id_11>",
+    "<extra_id_12>",
+    "<extra_id_13>",
+    "<extra_id_14>",
+    "<extra_id_15>",
+    "<extra_id_16>",
+    "<extra_id_17>",
+    "<extra_id_18>",
+    "<extra_id_19>",
+    "<extra_id_20>",
+    "<extra_id_21>",
+    "<extra_id_22>",
+    "<extra_id_23>",
+    "<extra_id_24>",
+    "<extra_id_25>",
+    "<extra_id_26>",
+    "<extra_id_27>",
+    "<extra_id_28>",
+    "<extra_id_29>",
+    "<extra_id_30>",
+    "<extra_id_31>",
+    "<extra_id_32>",
+    "<extra_id_33>",
+    "<extra_id_34>",
+    "<extra_id_35>",
+    "<extra_id_36>",
+    "<extra_id_37>",
+    "<extra_id_38>",
+    "<extra_id_39>",
+    "<extra_id_40>",
+    "<extra_id_41>",
+    "<extra_id_42>",
+    "<extra_id_43>",
+    "<extra_id_44>",
+    "<extra_id_45>",
+    "<extra_id_46>",
+    "<extra_id_47>",
+    "<extra_id_48>",
+    "<extra_id_49>",
+    "<extra_id_50>",
+    "<extra_id_51>",
+    "<extra_id_52>",
+    "<extra_id_53>",
+    "<extra_id_54>",
+    "<extra_id_55>",
+    "<extra_id_56>",
+    "<extra_id_57>",
+    "<extra_id_58>",
+    "<extra_id_59>",
+    "<extra_id_60>",
+    "<extra_id_61>",
+    "<extra_id_62>",
+    "<extra_id_63>",
+    "<extra_id_64>",
+    "<extra_id_65>",
+    "<extra_id_66>",
+    "<extra_id_67>",
+    "<extra_id_68>",
+    "<extra_id_69>",
+    "<extra_id_70>",
+    "<extra_id_71>",
+    "<extra_id_72>",
+    "<extra_id_73>",
+    "<extra_id_74>",
+    "<extra_id_75>",
+    "<extra_id_76>",
+    "<extra_id_77>",
+    "<extra_id_78>",
+    "<extra_id_79>",
+    "<extra_id_80>",
+    "<extra_id_81>",
+    "<extra_id_82>",
+    "<extra_id_83>",
+    "<extra_id_84>",
+    "<extra_id_85>",
+    "<extra_id_86>",
+    "<extra_id_87>",
+    "<extra_id_88>",
+    "<extra_id_89>",
+    "<extra_id_90>",
+    "<extra_id_91>",
+    "<extra_id_92>",
+    "<extra_id_93>",
+    "<extra_id_94>",
+    "<extra_id_95>",
+    "<extra_id_96>",
+    "<extra_id_97>",
+    "<extra_id_98>",
+    "<extra_id_99>"
+  ],
+  "clean_up_tokenization_spaces": true,
+  "eos_token": "</s>",
+  "extra_ids": 100,
+  "model_max_length": 512,
+  "pad_token": "<pad>",
+  "sp_model_kwargs": {},
+  "tokenizer_class": "T5Tokenizer",
+  "unk_token": "<unk>"
+}

fidmrg_answer.bm25.fidcqg-w/checkpoint-10000/trainer_state.json

@@ -0,0 +1,299 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.2597631645250693,
+  "eval_steps": 500,
+  "global_step": 10000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.5e-05,
+      "loss": 57.2732,
+      "step": 500
+    },
+    {
+      "epoch": 0.06,
+      "eval_loss": 51.029930114746094,
+      "eval_runtime": 3.1759,
+      "eval_samples_per_second": 31.487,
+      "eval_steps_per_second": 4.093,
+      "step": 500
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9e-05,
+      "loss": 56.0705,
+      "step": 1000
+    },
+    {
+      "epoch": 0.13,
+      "eval_loss": 49.898155212402344,
+      "eval_runtime": 4.4269,
+      "eval_samples_per_second": 22.589,
+      "eval_steps_per_second": 2.937,
+      "step": 1000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 8.5e-05,
+      "loss": 55.3144,
+      "step": 1500
+    },
+    {
+      "epoch": 0.19,
+      "eval_loss": 49.3874397277832,
+      "eval_runtime": 4.228,
+      "eval_samples_per_second": 23.652,
+      "eval_steps_per_second": 3.075,
+      "step": 1500
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8e-05,
+      "loss": 54.301,
+      "step": 2000
+    },
+    {
+      "epoch": 0.25,
+      "eval_loss": 48.694358825683594,
+      "eval_runtime": 3.1949,
+      "eval_samples_per_second": 31.299,
+      "eval_steps_per_second": 4.069,
+      "step": 2000
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 7.500000000000001e-05,
+      "loss": 54.8258,
+      "step": 2500
+    },
+    {
+      "epoch": 0.31,
+      "eval_loss": 49.120418548583984,
+      "eval_runtime": 4.0702,
+      "eval_samples_per_second": 24.569,
+      "eval_steps_per_second": 3.194,
+      "step": 2500
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 7e-05,
+      "loss": 55.0078,
+      "step": 3000
+    },
+    {
+      "epoch": 0.38,
+      "eval_loss": 48.479034423828125,
+      "eval_runtime": 3.9143,
+      "eval_samples_per_second": 25.547,
+      "eval_steps_per_second": 3.321,
+      "step": 3000
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 6.500000000000001e-05,
+      "loss": 55.0228,
+      "step": 3500
+    },
+    {
+      "epoch": 0.44,
+      "eval_loss": 48.023162841796875,
+      "eval_runtime": 3.9112,
+      "eval_samples_per_second": 25.568,
+      "eval_steps_per_second": 3.324,
+      "step": 3500
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 6e-05,
+      "loss": 53.2773,
+      "step": 4000
+    },
+    {
+      "epoch": 0.5,
+      "eval_loss": 47.419307708740234,
+      "eval_runtime": 4.7859,
+      "eval_samples_per_second": 20.895,
+      "eval_steps_per_second": 2.716,
+      "step": 4000
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.500000000000001e-05,
+      "loss": 54.9082,
+      "step": 4500
+    },
+    {
+      "epoch": 0.57,
+      "eval_loss": 46.841548919677734,
+      "eval_runtime": 4.8111,
+      "eval_samples_per_second": 20.785,
+      "eval_steps_per_second": 2.702,
+      "step": 4500
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 53.3684,
+      "step": 5000
+    },
+    {
+      "epoch": 0.63,
+      "eval_loss": 46.429256439208984,
+      "eval_runtime": 4.525,
+      "eval_samples_per_second": 22.099,
+      "eval_steps_per_second": 2.873,
+      "step": 5000
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.5e-05,
+      "loss": 54.4457,
+      "step": 5500
+    },
+    {
+      "epoch": 0.69,
+      "eval_loss": 46.04310607910156,
+      "eval_runtime": 4.6243,
+      "eval_samples_per_second": 21.625,
+      "eval_steps_per_second": 2.811,
+      "step": 5500
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4e-05,
+      "loss": 53.9792,
+      "step": 6000
+    },
+    {
+      "epoch": 0.76,
+      "eval_loss": 45.82972717285156,
+      "eval_runtime": 3.6612,
+      "eval_samples_per_second": 27.314,
+      "eval_steps_per_second": 3.551,
+      "step": 6000
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.5e-05,
+      "loss": 53.924,
+      "step": 6500
+    },
+    {
+      "epoch": 0.82,
+      "eval_loss": 45.8700065612793,
+      "eval_runtime": 4.1117,
+      "eval_samples_per_second": 24.321,
+      "eval_steps_per_second": 3.162,
+      "step": 6500
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3e-05,
+      "loss": 54.2944,
+      "step": 7000
+    },
+    {
+      "epoch": 0.88,
+      "eval_loss": 45.5869140625,
+      "eval_runtime": 4.5188,
+      "eval_samples_per_second": 22.13,
+      "eval_steps_per_second": 2.877,
+      "step": 7000
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.5e-05,
+      "loss": 53.9664,
+      "step": 7500
+    },
+    {
+      "epoch": 0.94,
+      "eval_loss": 45.52823257446289,
+      "eval_runtime": 4.5014,
+      "eval_samples_per_second": 22.215,
+      "eval_steps_per_second": 2.888,
+      "step": 7500
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2e-05,
+      "loss": 52.0975,
+      "step": 8000
+    },
+    {
+      "epoch": 1.01,
+      "eval_loss": 45.37757873535156,
+      "eval_runtime": 4.4906,
+      "eval_samples_per_second": 22.269,
+      "eval_steps_per_second": 2.895,
+      "step": 8000
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.5e-05,
+      "loss": 50.0576,
+      "step": 8500
+    },
+    {
+      "epoch": 1.07,
+      "eval_loss": 45.175968170166016,
+      "eval_runtime": 4.5314,
+      "eval_samples_per_second": 22.068,
+      "eval_steps_per_second": 2.869,
+      "step": 8500
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1e-05,
+      "loss": 50.6407,
+      "step": 9000
+    },
+    {
+      "epoch": 1.13,
+      "eval_loss": 45.02992630004883,
+      "eval_runtime": 4.196,
+      "eval_samples_per_second": 23.832,
+      "eval_steps_per_second": 3.098,
+      "step": 9000
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-06,
+      "loss": 50.3118,
+      "step": 9500
+    },
+    {
+      "epoch": 1.2,
+      "eval_loss": 44.964359283447266,
+      "eval_runtime": 4.7953,
+      "eval_samples_per_second": 20.854,
+      "eval_steps_per_second": 2.711,
+      "step": 9500
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.0,
+      "loss": 48.9218,
+      "step": 10000
+    },
+    {
+      "epoch": 1.26,
+      "eval_loss": 44.92063903808594,
+      "eval_runtime": 4.609,
+      "eval_samples_per_second": 21.697,
+      "eval_steps_per_second": 2.821,
+      "step": 10000
+    }
+  ],
+  "logging_steps": 500,
+  "max_steps": 10000,
+  "num_train_epochs": 2,
+  "save_steps": 5000,
+  "total_flos": 1.3694633480945664e+17,
+  "trial_name": null,
+  "trial_params": null
+}

fidmrg_answer.bm25.fidcqg-w/checkpoint-10000/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bd5596d549c770f38076b75ce0924f1801d992e81220f6d8bc270e00f65960db
+size 4472

fidmrg_answer.bm25.fidcqg-w/checkpoint-5000/config.json

@@ -0,0 +1,62 @@
+{
+  "_name_or_path": "google/flan-t5-base",
+  "architectures": [
+    "FiDT5"
+  ],
+  "classifier_dropout": 0.0,
+  "d_ff": 2048,
+  "d_kv": 64,
+  "d_model": 768,
+  "decoder_start_token_id": 0,
+  "dense_act_fn": "gelu_new",
+  "dropout_rate": 0.1,
+  "eos_token_id": 1,
+  "feed_forward_proj": "gated-gelu",
+  "initializer_factor": 1.0,
+  "is_encoder_decoder": true,
+  "is_gated_act": true,
+  "layer_norm_epsilon": 1e-06,
+  "model_type": "t5",
+  "n_positions": 512,
+  "num_decoder_layers": 12,
+  "num_heads": 12,
+  "num_layers": 12,
+  "output_past": true,
+  "pad_token_id": 0,
+  "relative_attention_max_distance": 128,
+  "relative_attention_num_buckets": 32,
+  "task_specific_params": {
+    "summarization": {
+      "early_stopping": true,
+      "length_penalty": 2.0,
+      "max_length": 200,
+      "min_length": 30,
+      "no_repeat_ngram_size": 3,
+      "num_beams": 4,
+      "prefix": "summarize: "
+    },
+    "translation_en_to_de": {
+      "early_stopping": true,
+      "max_length": 300,
+      "num_beams": 4,
+      "prefix": "translate English to German: "
+    },
+    "translation_en_to_fr": {
+      "early_stopping": true,
+      "max_length": 300,
+      "num_beams": 4,
+      "prefix": "translate English to French: "
+    },
+    "translation_en_to_ro": {
+      "early_stopping": true,
+      "max_length": 300,
+      "num_beams": 4,
+      "prefix": "translate English to Romanian: "
+    }
+  },
+  "tie_word_embeddings": false,
+  "torch_dtype": "float32",
+  "transformers_version": "4.33.1",
+  "use_cache": true,
+  "vocab_size": 32128
+}

fidmrg_answer.bm25.fidcqg-w/checkpoint-5000/generation_config.json

@@ -0,0 +1,7 @@
+{
+  "_from_model_config": true,
+  "decoder_start_token_id": 0,
+  "eos_token_id": 1,
+  "pad_token_id": 0,
+  "transformers_version": "4.33.1"
+}

fidmrg_answer.bm25.fidcqg-w/checkpoint-5000/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d20f60a3295b74a07b267d61c25c68b7d4aa25172ac5daf91ff43784aa899aa5
+size 1980864314

fidmrg_answer.bm25.fidcqg-w/checkpoint-5000/pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2f5f7db81aba54158bf048357fd142642374958f03cb6fb473a3b92cb6ad681e
+size 990411190

fidmrg_answer.bm25.fidcqg-w/checkpoint-5000/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:64e0d46eb43cded34db6d2c2e318774f14fbb9a206f56f7984381e69049a4151
+size 14244

fidmrg_answer.bm25.fidcqg-w/checkpoint-5000/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1a54271c874708bb67c02c881532d3a933ce5fa5776358051b2434fd912493cc
+size 1064

fidmrg_answer.bm25.fidcqg-w/checkpoint-5000/special_tokens_map.json

@@ -0,0 +1,107 @@
+{
+  "additional_special_tokens": [
+    "<extra_id_0>",
+    "<extra_id_1>",
+    "<extra_id_2>",
+    "<extra_id_3>",
+    "<extra_id_4>",
+    "<extra_id_5>",
+    "<extra_id_6>",
+    "<extra_id_7>",
+    "<extra_id_8>",
+    "<extra_id_9>",
+    "<extra_id_10>",
+    "<extra_id_11>",
+    "<extra_id_12>",
+    "<extra_id_13>",
+    "<extra_id_14>",
+    "<extra_id_15>",
+    "<extra_id_16>",
+    "<extra_id_17>",
+    "<extra_id_18>",
+    "<extra_id_19>",
+    "<extra_id_20>",
+    "<extra_id_21>",
+    "<extra_id_22>",
+    "<extra_id_23>",
+    "<extra_id_24>",
+    "<extra_id_25>",
+    "<extra_id_26>",
+    "<extra_id_27>",
+    "<extra_id_28>",
+    "<extra_id_29>",
+    "<extra_id_30>",
+    "<extra_id_31>",
+    "<extra_id_32>",
+    "<extra_id_33>",
+    "<extra_id_34>",
+    "<extra_id_35>",
+    "<extra_id_36>",
+    "<extra_id_37>",
+    "<extra_id_38>",
+    "<extra_id_39>",
+    "<extra_id_40>",
+    "<extra_id_41>",
+    "<extra_id_42>",
+    "<extra_id_43>",
+    "<extra_id_44>",
+    "<extra_id_45>",
+    "<extra_id_46>",
+    "<extra_id_47>",
+    "<extra_id_48>",
+    "<extra_id_49>",
+    "<extra_id_50>",
+    "<extra_id_51>",
+    "<extra_id_52>",
+    "<extra_id_53>",
+    "<extra_id_54>",
+    "<extra_id_55>",
+    "<extra_id_56>",
+    "<extra_id_57>",
+    "<extra_id_58>",
+    "<extra_id_59>",
+    "<extra_id_60>",
+    "<extra_id_61>",
+    "<extra_id_62>",
+    "<extra_id_63>",
+    "<extra_id_64>",
+    "<extra_id_65>",
+    "<extra_id_66>",
+    "<extra_id_67>",
+    "<extra_id_68>",
+    "<extra_id_69>",
+    "<extra_id_70>",
+    "<extra_id_71>",
+    "<extra_id_72>",
+    "<extra_id_73>",
+    "<extra_id_74>",
+    "<extra_id_75>",
+    "<extra_id_76>",
+    "<extra_id_77>",
+    "<extra_id_78>",
+    "<extra_id_79>",
+    "<extra_id_80>",
+    "<extra_id_81>",
+    "<extra_id_82>",
+    "<extra_id_83>",
+    "<extra_id_84>",
+    "<extra_id_85>",
+    "<extra_id_86>",
+    "<extra_id_87>",
+    "<extra_id_88>",
+    "<extra_id_89>",
+    "<extra_id_90>",
+    "<extra_id_91>",
+    "<extra_id_92>",
+    "<extra_id_93>",
+    "<extra_id_94>",
+    "<extra_id_95>",
+    "<extra_id_96>",
+    "<extra_id_97>",
+    "<extra_id_98>",
+    "<extra_id_99>"
+  ],
+  "eos_token": "</s>",
+  "pad_token": "<pad>",
+  "unk_token": "<unk>"
+}

fidmrg_answer.bm25.fidcqg-w/checkpoint-5000/spiece.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d60acb128cf7b7f2536e8f38a5b18a05535c9e14c7a355904270e15b0945ea86
+size 791656

fidmrg_answer.bm25.fidcqg-w/checkpoint-5000/tokenizer_config.json

@@ -0,0 +1,112 @@
+{
+  "additional_special_tokens": [
+    "<extra_id_0>",
+    "<extra_id_1>",
+    "<extra_id_2>",
+    "<extra_id_3>",
+    "<extra_id_4>",
+    "<extra_id_5>",
+    "<extra_id_6>",
+    "<extra_id_7>",
+    "<extra_id_8>",
+    "<extra_id_9>",
+    "<extra_id_10>",
+    "<extra_id_11>",
+    "<extra_id_12>",
+    "<extra_id_13>",
+    "<extra_id_14>",
+    "<extra_id_15>",
+    "<extra_id_16>",
+    "<extra_id_17>",
+    "<extra_id_18>",
+    "<extra_id_19>",
+    "<extra_id_20>",
+    "<extra_id_21>",
+    "<extra_id_22>",
+    "<extra_id_23>",
+    "<extra_id_24>",
+    "<extra_id_25>",
+    "<extra_id_26>",
+    "<extra_id_27>",
+    "<extra_id_28>",
+    "<extra_id_29>",
+    "<extra_id_30>",
+    "<extra_id_31>",
+    "<extra_id_32>",
+    "<extra_id_33>",
+    "<extra_id_34>",
+    "<extra_id_35>",
+    "<extra_id_36>",
+    "<extra_id_37>",
+    "<extra_id_38>",
+    "<extra_id_39>",
+    "<extra_id_40>",
+    "<extra_id_41>",
+    "<extra_id_42>",
+    "<extra_id_43>",
+    "<extra_id_44>",
+    "<extra_id_45>",
+    "<extra_id_46>",
+    "<extra_id_47>",
+    "<extra_id_48>",
+    "<extra_id_49>",
+    "<extra_id_50>",
+    "<extra_id_51>",
+    "<extra_id_52>",
+    "<extra_id_53>",
+    "<extra_id_54>",
+    "<extra_id_55>",
+    "<extra_id_56>",
+    "<extra_id_57>",
+    "<extra_id_58>",
+    "<extra_id_59>",
+    "<extra_id_60>",
+    "<extra_id_61>",
+    "<extra_id_62>",
+    "<extra_id_63>",
+    "<extra_id_64>",
+    "<extra_id_65>",
+    "<extra_id_66>",
+    "<extra_id_67>",
+    "<extra_id_68>",
+    "<extra_id_69>",
+    "<extra_id_70>",
+    "<extra_id_71>",
+    "<extra_id_72>",
+    "<extra_id_73>",
+    "<extra_id_74>",
+    "<extra_id_75>",
+    "<extra_id_76>",
+    "<extra_id_77>",
+    "<extra_id_78>",
+    "<extra_id_79>",
+    "<extra_id_80>",
+    "<extra_id_81>",
+    "<extra_id_82>",
+    "<extra_id_83>",
+    "<extra_id_84>",
+    "<extra_id_85>",
+    "<extra_id_86>",
+    "<extra_id_87>",
+    "<extra_id_88>",
+    "<extra_id_89>",
+    "<extra_id_90>",
+    "<extra_id_91>",
+    "<extra_id_92>",
+    "<extra_id_93>",
+    "<extra_id_94>",
+    "<extra_id_95>",
+    "<extra_id_96>",
+    "<extra_id_97>",
+    "<extra_id_98>",
+    "<extra_id_99>"
+  ],
+  "clean_up_tokenization_spaces": true,
+  "eos_token": "</s>",
+  "extra_ids": 100,
+  "model_max_length": 512,
+  "pad_token": "<pad>",
+  "sp_model_kwargs": {},
+  "tokenizer_class": "T5Tokenizer",
+  "unk_token": "<unk>"
+}

fidmrg_answer.bm25.fidcqg-w/checkpoint-5000/trainer_state.json

@@ -0,0 +1,159 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.6298815822625347,
+  "eval_steps": 500,
+  "global_step": 5000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.5e-05,
+      "loss": 57.2732,
+      "step": 500
+    },
+    {
+      "epoch": 0.06,
+      "eval_loss": 51.029930114746094,
+      "eval_runtime": 3.1759,
+      "eval_samples_per_second": 31.487,
+      "eval_steps_per_second": 4.093,
+      "step": 500
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9e-05,
+      "loss": 56.0705,
+      "step": 1000
+    },
+    {
+      "epoch": 0.13,
+      "eval_loss": 49.898155212402344,
+      "eval_runtime": 4.4269,
+      "eval_samples_per_second": 22.589,
+      "eval_steps_per_second": 2.937,
+      "step": 1000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 8.5e-05,
+      "loss": 55.3144,
+      "step": 1500
+    },
+    {
+      "epoch": 0.19,
+      "eval_loss": 49.3874397277832,
+      "eval_runtime": 4.228,
+      "eval_samples_per_second": 23.652,
+      "eval_steps_per_second": 3.075,
+      "step": 1500
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8e-05,
+      "loss": 54.301,
+      "step": 2000
+    },
+    {
+      "epoch": 0.25,
+      "eval_loss": 48.694358825683594,
+      "eval_runtime": 3.1949,
+      "eval_samples_per_second": 31.299,
+      "eval_steps_per_second": 4.069,
+      "step": 2000
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 7.500000000000001e-05,
+      "loss": 54.8258,
+      "step": 2500
+    },
+    {
+      "epoch": 0.31,
+      "eval_loss": 49.120418548583984,
+      "eval_runtime": 4.0702,
+      "eval_samples_per_second": 24.569,
+      "eval_steps_per_second": 3.194,
+      "step": 2500
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 7e-05,
+      "loss": 55.0078,
+      "step": 3000
+    },
+    {
+      "epoch": 0.38,
+      "eval_loss": 48.479034423828125,
+      "eval_runtime": 3.9143,
+      "eval_samples_per_second": 25.547,
+      "eval_steps_per_second": 3.321,
+      "step": 3000
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 6.500000000000001e-05,
+      "loss": 55.0228,
+      "step": 3500
+    },
+    {
+      "epoch": 0.44,
+      "eval_loss": 48.023162841796875,
+      "eval_runtime": 3.9112,
+      "eval_samples_per_second": 25.568,
+      "eval_steps_per_second": 3.324,
+      "step": 3500
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 6e-05,
+      "loss": 53.2773,
+      "step": 4000
+    },
+    {
+      "epoch": 0.5,
+      "eval_loss": 47.419307708740234,
+      "eval_runtime": 4.7859,
+      "eval_samples_per_second": 20.895,
+      "eval_steps_per_second": 2.716,
+      "step": 4000
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.500000000000001e-05,
+      "loss": 54.9082,
+      "step": 4500
+    },
+    {
+      "epoch": 0.57,
+      "eval_loss": 46.841548919677734,
+      "eval_runtime": 4.8111,
+      "eval_samples_per_second": 20.785,
+      "eval_steps_per_second": 2.702,
+      "step": 4500
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 53.3684,
+      "step": 5000
+    },
+    {
+      "epoch": 0.63,
+      "eval_loss": 46.429256439208984,
+      "eval_runtime": 4.525,
+      "eval_samples_per_second": 22.099,
+      "eval_steps_per_second": 2.873,
+      "step": 5000
+    }
+  ],
+  "logging_steps": 500,
+  "max_steps": 10000,
+  "num_train_epochs": 2,
+  "save_steps": 5000,
+  "total_flos": 6.84757352448e+16,
+  "trial_name": null,
+  "trial_params": null
+}

fidmrg_answer.bm25.fidcqg-w/checkpoint-5000/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bd5596d549c770f38076b75ce0924f1801d992e81220f6d8bc270e00f65960db
+size 4472

fidmrg_enumerated.bm25.fidcqg-w/checkpoint-10000/config.json

@@ -0,0 +1,62 @@
+{
+  "_name_or_path": "google/flan-t5-base",
+  "architectures": [
+    "FiDT5"
+  ],
+  "classifier_dropout": 0.0,
+  "d_ff": 2048,
+  "d_kv": 64,
+  "d_model": 768,
+  "decoder_start_token_id": 0,
+  "dense_act_fn": "gelu_new",
+  "dropout_rate": 0.1,
+  "eos_token_id": 1,
+  "feed_forward_proj": "gated-gelu",
+  "initializer_factor": 1.0,
+  "is_encoder_decoder": true,
+  "is_gated_act": true,
+  "layer_norm_epsilon": 1e-06,
+  "model_type": "t5",
+  "n_positions": 512,
+  "num_decoder_layers": 12,
+  "num_heads": 12,
+  "num_layers": 12,
+  "output_past": true,
+  "pad_token_id": 0,
+  "relative_attention_max_distance": 128,
+  "relative_attention_num_buckets": 32,
+  "task_specific_params": {
+    "summarization": {
+      "early_stopping": true,
+      "length_penalty": 2.0,
+      "max_length": 200,
+      "min_length": 30,
+      "no_repeat_ngram_size": 3,
+      "num_beams": 4,
+      "prefix": "summarize: "
+    },
+    "translation_en_to_de": {
+      "early_stopping": true,
+      "max_length": 300,
+      "num_beams": 4,
+      "prefix": "translate English to German: "
+    },
+    "translation_en_to_fr": {
+      "early_stopping": true,
+      "max_length": 300,
+      "num_beams": 4,
+      "prefix": "translate English to French: "
+    },
+    "translation_en_to_ro": {
+      "early_stopping": true,
+      "max_length": 300,
+      "num_beams": 4,
+      "prefix": "translate English to Romanian: "
+    }
+  },
+  "tie_word_embeddings": false,
+  "torch_dtype": "float32",
+  "transformers_version": "4.33.1",
+  "use_cache": true,
+  "vocab_size": 32128
+}

fidmrg_enumerated.bm25.fidcqg-w/checkpoint-10000/generation_config.json

@@ -0,0 +1,7 @@
+{
+  "_from_model_config": true,
+  "decoder_start_token_id": 0,
+  "eos_token_id": 1,
+  "pad_token_id": 0,
+  "transformers_version": "4.33.1"
+}

fidmrg_enumerated.bm25.fidcqg-w/checkpoint-10000/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bd17bcfe3aacca1b46ca6bcc5b27e83d895f034b414db1a22ea522ad9d2edb60
+size 1980864314

fidmrg_enumerated.bm25.fidcqg-w/checkpoint-10000/pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:32583cd56b50201a41a61cc96ab9b90f5657883caeece052e9789534a13fbf4a
+size 990411190

fidmrg_enumerated.bm25.fidcqg-w/checkpoint-10000/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:603aba05bfe4e59fa0c1c9b0281c6f463fe8d68cd35d50ee6a73070cbba623d3
+size 14244

fidmrg_enumerated.bm25.fidcqg-w/checkpoint-10000/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:07356cb668b8efc440ddb2f7f81f7da3152060e81d0229922fa3b30df3313f82
+size 1064

fidmrg_enumerated.bm25.fidcqg-w/checkpoint-10000/special_tokens_map.json

@@ -0,0 +1,107 @@
+{
+  "additional_special_tokens": [
+    "<extra_id_0>",
+    "<extra_id_1>",
+    "<extra_id_2>",
+    "<extra_id_3>",
+    "<extra_id_4>",
+    "<extra_id_5>",
+    "<extra_id_6>",
+    "<extra_id_7>",
+    "<extra_id_8>",
+    "<extra_id_9>",
+    "<extra_id_10>",
+    "<extra_id_11>",
+    "<extra_id_12>",
+    "<extra_id_13>",
+    "<extra_id_14>",
+    "<extra_id_15>",
+    "<extra_id_16>",
+    "<extra_id_17>",
+    "<extra_id_18>",
+    "<extra_id_19>",
+    "<extra_id_20>",
+    "<extra_id_21>",
+    "<extra_id_22>",
+    "<extra_id_23>",
+    "<extra_id_24>",
+    "<extra_id_25>",
+    "<extra_id_26>",
+    "<extra_id_27>",
+    "<extra_id_28>",
+    "<extra_id_29>",
+    "<extra_id_30>",
+    "<extra_id_31>",
+    "<extra_id_32>",
+    "<extra_id_33>",
+    "<extra_id_34>",
+    "<extra_id_35>",
+    "<extra_id_36>",
+    "<extra_id_37>",
+    "<extra_id_38>",
+    "<extra_id_39>",
+    "<extra_id_40>",
+    "<extra_id_41>",
+    "<extra_id_42>",
+    "<extra_id_43>",
+    "<extra_id_44>",
+    "<extra_id_45>",
+    "<extra_id_46>",
+    "<extra_id_47>",
+    "<extra_id_48>",
+    "<extra_id_49>",
+    "<extra_id_50>",
+    "<extra_id_51>",
+    "<extra_id_52>",
+    "<extra_id_53>",
+    "<extra_id_54>",
+    "<extra_id_55>",
+    "<extra_id_56>",
+    "<extra_id_57>",
+    "<extra_id_58>",
+    "<extra_id_59>",
+    "<extra_id_60>",
+    "<extra_id_61>",
+    "<extra_id_62>",
+    "<extra_id_63>",
+    "<extra_id_64>",
+    "<extra_id_65>",
+    "<extra_id_66>",
+    "<extra_id_67>",
+    "<extra_id_68>",
+    "<extra_id_69>",
+    "<extra_id_70>",
+    "<extra_id_71>",
+    "<extra_id_72>",
+    "<extra_id_73>",
+    "<extra_id_74>",
+    "<extra_id_75>",
+    "<extra_id_76>",
+    "<extra_id_77>",
+    "<extra_id_78>",
+    "<extra_id_79>",
+    "<extra_id_80>",
+    "<extra_id_81>",
+    "<extra_id_82>",
+    "<extra_id_83>",
+    "<extra_id_84>",
+    "<extra_id_85>",
+    "<extra_id_86>",
+    "<extra_id_87>",
+    "<extra_id_88>",
+    "<extra_id_89>",
+    "<extra_id_90>",
+    "<extra_id_91>",
+    "<extra_id_92>",
+    "<extra_id_93>",
+    "<extra_id_94>",
+    "<extra_id_95>",
+    "<extra_id_96>",
+    "<extra_id_97>",
+    "<extra_id_98>",
+    "<extra_id_99>"
+  ],
+  "eos_token": "</s>",
+  "pad_token": "<pad>",
+  "unk_token": "<unk>"
+}

fidmrg_enumerated.bm25.fidcqg-w/checkpoint-10000/spiece.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d60acb128cf7b7f2536e8f38a5b18a05535c9e14c7a355904270e15b0945ea86
+size 791656

fidmrg_enumerated.bm25.fidcqg-w/checkpoint-10000/tokenizer_config.json

@@ -0,0 +1,112 @@
+{
+  "additional_special_tokens": [
+    "<extra_id_0>",
+    "<extra_id_1>",
+    "<extra_id_2>",
+    "<extra_id_3>",
+    "<extra_id_4>",
+    "<extra_id_5>",
+    "<extra_id_6>",
+    "<extra_id_7>",
+    "<extra_id_8>",
+    "<extra_id_9>",
+    "<extra_id_10>",
+    "<extra_id_11>",
+    "<extra_id_12>",
+    "<extra_id_13>",
+    "<extra_id_14>",
+    "<extra_id_15>",
+    "<extra_id_16>",
+    "<extra_id_17>",
+    "<extra_id_18>",
+    "<extra_id_19>",
+    "<extra_id_20>",
+    "<extra_id_21>",
+    "<extra_id_22>",
+    "<extra_id_23>",
+    "<extra_id_24>",
+    "<extra_id_25>",
+    "<extra_id_26>",
+    "<extra_id_27>",
+    "<extra_id_28>",
+    "<extra_id_29>",
+    "<extra_id_30>",
+    "<extra_id_31>",
+    "<extra_id_32>",
+    "<extra_id_33>",
+    "<extra_id_34>",
+    "<extra_id_35>",
+    "<extra_id_36>",
+    "<extra_id_37>",
+    "<extra_id_38>",
+    "<extra_id_39>",
+    "<extra_id_40>",
+    "<extra_id_41>",
+    "<extra_id_42>",
+    "<extra_id_43>",
+    "<extra_id_44>",
+    "<extra_id_45>",
+    "<extra_id_46>",
+    "<extra_id_47>",
+    "<extra_id_48>",
+    "<extra_id_49>",
+    "<extra_id_50>",
+    "<extra_id_51>",
+    "<extra_id_52>",
+    "<extra_id_53>",
+    "<extra_id_54>",
+    "<extra_id_55>",
+    "<extra_id_56>",
+    "<extra_id_57>",
+    "<extra_id_58>",
+    "<extra_id_59>",
+    "<extra_id_60>",
+    "<extra_id_61>",
+    "<extra_id_62>",
+    "<extra_id_63>",
+    "<extra_id_64>",
+    "<extra_id_65>",
+    "<extra_id_66>",
+    "<extra_id_67>",
+    "<extra_id_68>",
+    "<extra_id_69>",
+    "<extra_id_70>",
+    "<extra_id_71>",
+    "<extra_id_72>",
+    "<extra_id_73>",
+    "<extra_id_74>",
+    "<extra_id_75>",
+    "<extra_id_76>",
+    "<extra_id_77>",
+    "<extra_id_78>",
+    "<extra_id_79>",
+    "<extra_id_80>",
+    "<extra_id_81>",
+    "<extra_id_82>",
+    "<extra_id_83>",
+    "<extra_id_84>",
+    "<extra_id_85>",
+    "<extra_id_86>",
+    "<extra_id_87>",
+    "<extra_id_88>",
+    "<extra_id_89>",
+    "<extra_id_90>",
+    "<extra_id_91>",
+    "<extra_id_92>",
+    "<extra_id_93>",
+    "<extra_id_94>",
+    "<extra_id_95>",
+    "<extra_id_96>",
+    "<extra_id_97>",
+    "<extra_id_98>",
+    "<extra_id_99>"
+  ],
+  "clean_up_tokenization_spaces": true,
+  "eos_token": "</s>",
+  "extra_ids": 100,
+  "model_max_length": 512,
+  "pad_token": "<pad>",
+  "sp_model_kwargs": {},
+  "tokenizer_class": "T5Tokenizer",
+  "unk_token": "<unk>"
+}

fidmrg_enumerated.bm25.fidcqg-w/checkpoint-10000/trainer_state.json

@@ -0,0 +1,299 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.6298815822625347,
+  "eval_steps": 500,
+  "global_step": 10000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.5e-05,
+      "loss": 37.4397,
+      "step": 500
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 31.211898803710938,
+      "eval_runtime": 6.944,
+      "eval_samples_per_second": 14.401,
+      "eval_steps_per_second": 3.6,
+      "step": 500
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9e-05,
+      "loss": 33.7011,
+      "step": 1000
+    },
+    {
+      "epoch": 0.06,
+      "eval_loss": 30.628690719604492,
+      "eval_runtime": 8.8362,
+      "eval_samples_per_second": 11.317,
+      "eval_steps_per_second": 2.829,
+      "step": 1000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 8.5e-05,
+      "loss": 34.3949,
+      "step": 1500
+    },
+    {
+      "epoch": 0.09,
+      "eval_loss": 30.2351016998291,
+      "eval_runtime": 9.7911,
+      "eval_samples_per_second": 10.213,
+      "eval_steps_per_second": 2.553,
+      "step": 1500
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 8e-05,
+      "loss": 33.7534,
+      "step": 2000
+    },
+    {
+      "epoch": 0.13,
+      "eval_loss": 30.151206970214844,
+      "eval_runtime": 9.0893,
+      "eval_samples_per_second": 11.002,
+      "eval_steps_per_second": 2.75,
+      "step": 2000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 7.500000000000001e-05,
+      "loss": 33.9418,
+      "step": 2500
+    },
+    {
+      "epoch": 0.16,
+      "eval_loss": 30.117929458618164,
+      "eval_runtime": 7.9747,
+      "eval_samples_per_second": 12.54,
+      "eval_steps_per_second": 3.135,
+      "step": 2500
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 7e-05,
+      "loss": 33.3573,
+      "step": 3000
+    },
+    {
+      "epoch": 0.19,
+      "eval_loss": 29.591096878051758,
+      "eval_runtime": 7.5459,
+      "eval_samples_per_second": 13.252,
+      "eval_steps_per_second": 3.313,
+      "step": 3000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 6.500000000000001e-05,
+      "loss": 33.3372,
+      "step": 3500
+    },
+    {
+      "epoch": 0.22,
+      "eval_loss": 29.37584686279297,
+      "eval_runtime": 9.1309,
+      "eval_samples_per_second": 10.952,
+      "eval_steps_per_second": 2.738,
+      "step": 3500
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 6e-05,
+      "loss": 32.5095,
+      "step": 4000
+    },
+    {
+      "epoch": 0.25,
+      "eval_loss": 29.21722412109375,
+      "eval_runtime": 8.935,
+      "eval_samples_per_second": 11.192,
+      "eval_steps_per_second": 2.798,
+      "step": 4000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.500000000000001e-05,
+      "loss": 32.6192,
+      "step": 4500
+    },
+    {
+      "epoch": 0.28,
+      "eval_loss": 28.85555648803711,
+      "eval_runtime": 8.0464,
+      "eval_samples_per_second": 12.428,
+      "eval_steps_per_second": 3.107,
+      "step": 4500
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 32.902,
+      "step": 5000
+    },
+    {
+      "epoch": 0.31,
+      "eval_loss": 28.798521041870117,
+      "eval_runtime": 8.3773,
+      "eval_samples_per_second": 11.937,
+      "eval_steps_per_second": 2.984,
+      "step": 5000
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.5e-05,
+      "loss": 33.2756,
+      "step": 5500
+    },
+    {
+      "epoch": 0.35,
+      "eval_loss": 28.811630249023438,
+      "eval_runtime": 8.4098,
+      "eval_samples_per_second": 11.891,
+      "eval_steps_per_second": 2.973,
+      "step": 5500
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4e-05,
+      "loss": 32.3762,
+      "step": 6000
+    },
+    {
+      "epoch": 0.38,
+      "eval_loss": 28.51260757446289,
+      "eval_runtime": 8.6441,
+      "eval_samples_per_second": 11.569,
+      "eval_steps_per_second": 2.892,
+      "step": 6000
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 3.5e-05,
+      "loss": 32.5452,
+      "step": 6500
+    },
+    {
+      "epoch": 0.41,
+      "eval_loss": 28.600557327270508,
+      "eval_runtime": 7.2739,
+      "eval_samples_per_second": 13.748,
+      "eval_steps_per_second": 3.437,
+      "step": 6500
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 3e-05,
+      "loss": 33.0182,
+      "step": 7000
+    },
+    {
+      "epoch": 0.44,
+      "eval_loss": 28.495351791381836,
+      "eval_runtime": 6.2937,
+      "eval_samples_per_second": 15.889,
+      "eval_steps_per_second": 3.972,
+      "step": 7000
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.5e-05,
+      "loss": 31.8452,
+      "step": 7500
+    },
+    {
+      "epoch": 0.47,
+      "eval_loss": 28.361299514770508,
+      "eval_runtime": 8.0596,
+      "eval_samples_per_second": 12.407,
+      "eval_steps_per_second": 3.102,
+      "step": 7500
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2e-05,
+      "loss": 32.0009,
+      "step": 8000
+    },
+    {
+      "epoch": 0.5,
+      "eval_loss": 28.153642654418945,
+      "eval_runtime": 6.3145,
+      "eval_samples_per_second": 15.837,
+      "eval_steps_per_second": 3.959,
+      "step": 8000
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.5e-05,
+      "loss": 33.4938,
+      "step": 8500
+    },
+    {
+      "epoch": 0.54,
+      "eval_loss": 28.087358474731445,
+      "eval_runtime": 6.3508,
+      "eval_samples_per_second": 15.746,
+      "eval_steps_per_second": 3.937,
+      "step": 8500
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1e-05,
+      "loss": 31.9753,
+      "step": 9000
+    },
+    {
+      "epoch": 0.57,
+      "eval_loss": 28.113698959350586,
+      "eval_runtime": 9.8066,
+      "eval_samples_per_second": 10.197,
+      "eval_steps_per_second": 2.549,
+      "step": 9000
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-06,
+      "loss": 31.9571,
+      "step": 9500
+    },
+    {
+      "epoch": 0.6,
+      "eval_loss": 28.042316436767578,
+      "eval_runtime": 9.4344,
+      "eval_samples_per_second": 10.6,
+      "eval_steps_per_second": 2.65,
+      "step": 9500
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0,
+      "loss": 31.8931,
+      "step": 10000
+    },
+    {
+      "epoch": 0.63,
+      "eval_loss": 28.017412185668945,
+      "eval_runtime": 9.3747,
+      "eval_samples_per_second": 10.667,
+      "eval_steps_per_second": 2.667,
+      "step": 10000
+    }
+  ],
+  "logging_steps": 500,
+  "max_steps": 10000,
+  "num_train_epochs": 1,
+  "save_steps": 5000,
+  "total_flos": 1.369514704896e+17,
+  "trial_name": null,
+  "trial_params": null
+}

fidmrg_enumerated.bm25.fidcqg-w/checkpoint-10000/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6788bb6ce57287043165ad1f0e3348866b7c4ba15f6c60397df9721eb75d80a3
+size 4472

fidmrg_enumerated.bm25.fidcqg-w/checkpoint-5000/config.json

@@ -0,0 +1,62 @@
+{
+  "_name_or_path": "google/flan-t5-base",
+  "architectures": [
+    "FiDT5"
+  ],
+  "classifier_dropout": 0.0,
+  "d_ff": 2048,
+  "d_kv": 64,
+  "d_model": 768,
+  "decoder_start_token_id": 0,
+  "dense_act_fn": "gelu_new",
+  "dropout_rate": 0.1,
+  "eos_token_id": 1,
+  "feed_forward_proj": "gated-gelu",
+  "initializer_factor": 1.0,
+  "is_encoder_decoder": true,
+  "is_gated_act": true,
+  "layer_norm_epsilon": 1e-06,
+  "model_type": "t5",
+  "n_positions": 512,
+  "num_decoder_layers": 12,
+  "num_heads": 12,
+  "num_layers": 12,
+  "output_past": true,
+  "pad_token_id": 0,
+  "relative_attention_max_distance": 128,
+  "relative_attention_num_buckets": 32,
+  "task_specific_params": {
+    "summarization": {
+      "early_stopping": true,
+      "length_penalty": 2.0,
+      "max_length": 200,
+      "min_length": 30,
+      "no_repeat_ngram_size": 3,
+      "num_beams": 4,
+      "prefix": "summarize: "
+    },
+    "translation_en_to_de": {
+      "early_stopping": true,
+      "max_length": 300,
+      "num_beams": 4,
+      "prefix": "translate English to German: "
+    },
+    "translation_en_to_fr": {
+      "early_stopping": true,
+      "max_length": 300,
+      "num_beams": 4,
+      "prefix": "translate English to French: "
+    },
+    "translation_en_to_ro": {
+      "early_stopping": true,
+      "max_length": 300,
+      "num_beams": 4,
+      "prefix": "translate English to Romanian: "
+    }
+  },
+  "tie_word_embeddings": false,
+  "torch_dtype": "float32",
+  "transformers_version": "4.33.1",
+  "use_cache": true,
+  "vocab_size": 32128
+}

fidmrg_enumerated.bm25.fidcqg-w/checkpoint-5000/generation_config.json

@@ -0,0 +1,7 @@
+{
+  "_from_model_config": true,
+  "decoder_start_token_id": 0,
+  "eos_token_id": 1,
+  "pad_token_id": 0,
+  "transformers_version": "4.33.1"
+}

fidmrg_enumerated.bm25.fidcqg-w/checkpoint-5000/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a7877738917a82a130f1ca62fb594f7bf0f13c8d687263977828ff69018c7f93
+size 1980864314

fidmrg_enumerated.bm25.fidcqg-w/checkpoint-5000/pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0c25b2e6e086b34f050d9d56521d164ba99b3e7707dbd655cf6e4af8ef02d2d4
+size 990411190

fidmrg_enumerated.bm25.fidcqg-w/checkpoint-5000/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:01d4f44fd881db9d73886c213b534385398ac892b510ce9a9c6d21385fee986f
+size 14244

fidmrg_enumerated.bm25.fidcqg-w/checkpoint-5000/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1a54271c874708bb67c02c881532d3a933ce5fa5776358051b2434fd912493cc
+size 1064

fidmrg_enumerated.bm25.fidcqg-w/checkpoint-5000/special_tokens_map.json

@@ -0,0 +1,107 @@
+{
+  "additional_special_tokens": [
+    "<extra_id_0>",
+    "<extra_id_1>",
+    "<extra_id_2>",
+    "<extra_id_3>",
+    "<extra_id_4>",
+    "<extra_id_5>",
+    "<extra_id_6>",
+    "<extra_id_7>",
+    "<extra_id_8>",
+    "<extra_id_9>",
+    "<extra_id_10>",
+    "<extra_id_11>",
+    "<extra_id_12>",
+    "<extra_id_13>",
+    "<extra_id_14>",
+    "<extra_id_15>",
+    "<extra_id_16>",
+    "<extra_id_17>",
+    "<extra_id_18>",
+    "<extra_id_19>",
+    "<extra_id_20>",
+    "<extra_id_21>",
+    "<extra_id_22>",
+    "<extra_id_23>",
+    "<extra_id_24>",
+    "<extra_id_25>",
+    "<extra_id_26>",
+    "<extra_id_27>",
+    "<extra_id_28>",
+    "<extra_id_29>",
+    "<extra_id_30>",
+    "<extra_id_31>",
+    "<extra_id_32>",
+    "<extra_id_33>",
+    "<extra_id_34>",
+    "<extra_id_35>",
+    "<extra_id_36>",
+    "<extra_id_37>",
+    "<extra_id_38>",
+    "<extra_id_39>",
+    "<extra_id_40>",
+    "<extra_id_41>",
+    "<extra_id_42>",
+    "<extra_id_43>",
+    "<extra_id_44>",
+    "<extra_id_45>",
+    "<extra_id_46>",
+    "<extra_id_47>",
+    "<extra_id_48>",
+    "<extra_id_49>",
+    "<extra_id_50>",
+    "<extra_id_51>",
+    "<extra_id_52>",
+    "<extra_id_53>",
+    "<extra_id_54>",
+    "<extra_id_55>",
+    "<extra_id_56>",
+    "<extra_id_57>",
+    "<extra_id_58>",
+    "<extra_id_59>",
+    "<extra_id_60>",
+    "<extra_id_61>",
+    "<extra_id_62>",
+    "<extra_id_63>",
+    "<extra_id_64>",
+    "<extra_id_65>",
+    "<extra_id_66>",
+    "<extra_id_67>",
+    "<extra_id_68>",
+    "<extra_id_69>",
+    "<extra_id_70>",
+    "<extra_id_71>",
+    "<extra_id_72>",
+    "<extra_id_73>",
+    "<extra_id_74>",
+    "<extra_id_75>",
+    "<extra_id_76>",
+    "<extra_id_77>",
+    "<extra_id_78>",
+    "<extra_id_79>",
+    "<extra_id_80>",
+    "<extra_id_81>",
+    "<extra_id_82>",
+    "<extra_id_83>",
+    "<extra_id_84>",
+    "<extra_id_85>",
+    "<extra_id_86>",
+    "<extra_id_87>",
+    "<extra_id_88>",
+    "<extra_id_89>",
+    "<extra_id_90>",
+    "<extra_id_91>",
+    "<extra_id_92>",
+    "<extra_id_93>",
+    "<extra_id_94>",
+    "<extra_id_95>",
+    "<extra_id_96>",
+    "<extra_id_97>",
+    "<extra_id_98>",
+    "<extra_id_99>"
+  ],
+  "eos_token": "</s>",
+  "pad_token": "<pad>",
+  "unk_token": "<unk>"
+}

fidmrg_enumerated.bm25.fidcqg-w/checkpoint-5000/spiece.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d60acb128cf7b7f2536e8f38a5b18a05535c9e14c7a355904270e15b0945ea86
+size 791656

fidmrg_enumerated.bm25.fidcqg-w/checkpoint-5000/tokenizer_config.json

@@ -0,0 +1,112 @@
+{
+  "additional_special_tokens": [
+    "<extra_id_0>",
+    "<extra_id_1>",
+    "<extra_id_2>",
+    "<extra_id_3>",
+    "<extra_id_4>",
+    "<extra_id_5>",
+    "<extra_id_6>",
+    "<extra_id_7>",
+    "<extra_id_8>",
+    "<extra_id_9>",
+    "<extra_id_10>",
+    "<extra_id_11>",
+    "<extra_id_12>",
+    "<extra_id_13>",
+    "<extra_id_14>",
+    "<extra_id_15>",
+    "<extra_id_16>",
+    "<extra_id_17>",
+    "<extra_id_18>",
+    "<extra_id_19>",
+    "<extra_id_20>",
+    "<extra_id_21>",
+    "<extra_id_22>",
+    "<extra_id_23>",
+    "<extra_id_24>",
+    "<extra_id_25>",
+    "<extra_id_26>",
+    "<extra_id_27>",
+    "<extra_id_28>",
+    "<extra_id_29>",
+    "<extra_id_30>",
+    "<extra_id_31>",
+    "<extra_id_32>",
+    "<extra_id_33>",
+    "<extra_id_34>",
+    "<extra_id_35>",
+    "<extra_id_36>",
+    "<extra_id_37>",
+    "<extra_id_38>",
+    "<extra_id_39>",
+    "<extra_id_40>",
+    "<extra_id_41>",
+    "<extra_id_42>",
+    "<extra_id_43>",
+    "<extra_id_44>",
+    "<extra_id_45>",
+    "<extra_id_46>",
+    "<extra_id_47>",
+    "<extra_id_48>",
+    "<extra_id_49>",
+    "<extra_id_50>",
+    "<extra_id_51>",
+    "<extra_id_52>",
+    "<extra_id_53>",
+    "<extra_id_54>",
+    "<extra_id_55>",
+    "<extra_id_56>",
+    "<extra_id_57>",
+    "<extra_id_58>",
+    "<extra_id_59>",
+    "<extra_id_60>",
+    "<extra_id_61>",
+    "<extra_id_62>",
+    "<extra_id_63>",
+    "<extra_id_64>",
+    "<extra_id_65>",
+    "<extra_id_66>",
+    "<extra_id_67>",
+    "<extra_id_68>",
+    "<extra_id_69>",
+    "<extra_id_70>",
+    "<extra_id_71>",
+    "<extra_id_72>",
+    "<extra_id_73>",
+    "<extra_id_74>",
+    "<extra_id_75>",
+    "<extra_id_76>",
+    "<extra_id_77>",
+    "<extra_id_78>",
+    "<extra_id_79>",
+    "<extra_id_80>",
+    "<extra_id_81>",
+    "<extra_id_82>",
+    "<extra_id_83>",
+    "<extra_id_84>",
+    "<extra_id_85>",
+    "<extra_id_86>",
+    "<extra_id_87>",
+    "<extra_id_88>",
+    "<extra_id_89>",
+    "<extra_id_90>",
+    "<extra_id_91>",
+    "<extra_id_92>",
+    "<extra_id_93>",
+    "<extra_id_94>",
+    "<extra_id_95>",
+    "<extra_id_96>",
+    "<extra_id_97>",
+    "<extra_id_98>",
+    "<extra_id_99>"
+  ],
+  "clean_up_tokenization_spaces": true,
+  "eos_token": "</s>",
+  "extra_ids": 100,
+  "model_max_length": 512,
+  "pad_token": "<pad>",
+  "sp_model_kwargs": {},
+  "tokenizer_class": "T5Tokenizer",
+  "unk_token": "<unk>"
+}

fidmrg_enumerated.bm25.fidcqg-w/checkpoint-5000/trainer_state.json

@@ -0,0 +1,159 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.31494079113126733,
+  "eval_steps": 500,
+  "global_step": 5000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.5e-05,
+      "loss": 37.4397,
+      "step": 500
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 31.211898803710938,
+      "eval_runtime": 6.944,
+      "eval_samples_per_second": 14.401,
+      "eval_steps_per_second": 3.6,
+      "step": 500
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9e-05,
+      "loss": 33.7011,
+      "step": 1000
+    },
+    {
+      "epoch": 0.06,
+      "eval_loss": 30.628690719604492,
+      "eval_runtime": 8.8362,
+      "eval_samples_per_second": 11.317,
+      "eval_steps_per_second": 2.829,
+      "step": 1000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 8.5e-05,
+      "loss": 34.3949,
+      "step": 1500
+    },
+    {
+      "epoch": 0.09,
+      "eval_loss": 30.2351016998291,
+      "eval_runtime": 9.7911,
+      "eval_samples_per_second": 10.213,
+      "eval_steps_per_second": 2.553,
+      "step": 1500
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 8e-05,
+      "loss": 33.7534,
+      "step": 2000
+    },
+    {
+      "epoch": 0.13,
+      "eval_loss": 30.151206970214844,
+      "eval_runtime": 9.0893,
+      "eval_samples_per_second": 11.002,
+      "eval_steps_per_second": 2.75,
+      "step": 2000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 7.500000000000001e-05,
+      "loss": 33.9418,
+      "step": 2500
+    },
+    {
+      "epoch": 0.16,
+      "eval_loss": 30.117929458618164,
+      "eval_runtime": 7.9747,
+      "eval_samples_per_second": 12.54,
+      "eval_steps_per_second": 3.135,
+      "step": 2500
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 7e-05,
+      "loss": 33.3573,
+      "step": 3000
+    },
+    {
+      "epoch": 0.19,
+      "eval_loss": 29.591096878051758,
+      "eval_runtime": 7.5459,
+      "eval_samples_per_second": 13.252,
+      "eval_steps_per_second": 3.313,
+      "step": 3000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 6.500000000000001e-05,
+      "loss": 33.3372,
+      "step": 3500
+    },
+    {
+      "epoch": 0.22,
+      "eval_loss": 29.37584686279297,
+      "eval_runtime": 9.1309,
+      "eval_samples_per_second": 10.952,
+      "eval_steps_per_second": 2.738,
+      "step": 3500
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 6e-05,
+      "loss": 32.5095,
+      "step": 4000
+    },
+    {
+      "epoch": 0.25,
+      "eval_loss": 29.21722412109375,
+      "eval_runtime": 8.935,
+      "eval_samples_per_second": 11.192,
+      "eval_steps_per_second": 2.798,
+      "step": 4000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.500000000000001e-05,
+      "loss": 32.6192,
+      "step": 4500
+    },
+    {
+      "epoch": 0.28,
+      "eval_loss": 28.85555648803711,
+      "eval_runtime": 8.0464,
+      "eval_samples_per_second": 12.428,
+      "eval_steps_per_second": 3.107,
+      "step": 4500
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 32.902,
+      "step": 5000
+    },
+    {
+      "epoch": 0.31,
+      "eval_loss": 28.798521041870117,
+      "eval_runtime": 8.3773,
+      "eval_samples_per_second": 11.937,
+      "eval_steps_per_second": 2.984,
+      "step": 5000
+    }
+  ],
+  "logging_steps": 500,
+  "max_steps": 10000,
+  "num_train_epochs": 1,
+  "save_steps": 5000,
+  "total_flos": 6.84757352448e+16,
+  "trial_name": null,
+  "trial_params": null
+}

fidmrg_enumerated.bm25.fidcqg-w/checkpoint-5000/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6788bb6ce57287043165ad1f0e3348866b7c4ba15f6c60397df9721eb75d80a3
+size 4472

fidmrg_random.bm25.fidcqg-w/checkpoint-10000/config.json

@@ -0,0 +1,62 @@
+{
+  "_name_or_path": "google/flan-t5-base",
+  "architectures": [
+    "FiDT5"
+  ],
+  "classifier_dropout": 0.0,
+  "d_ff": 2048,
+  "d_kv": 64,
+  "d_model": 768,
+  "decoder_start_token_id": 0,
+  "dense_act_fn": "gelu_new",
+  "dropout_rate": 0.1,
+  "eos_token_id": 1,
+  "feed_forward_proj": "gated-gelu",
+  "initializer_factor": 1.0,
+  "is_encoder_decoder": true,
+  "is_gated_act": true,
+  "layer_norm_epsilon": 1e-06,
+  "model_type": "t5",
+  "n_positions": 512,
+  "num_decoder_layers": 12,
+  "num_heads": 12,
+  "num_layers": 12,
+  "output_past": true,
+  "pad_token_id": 0,
+  "relative_attention_max_distance": 128,
+  "relative_attention_num_buckets": 32,
+  "task_specific_params": {
+    "summarization": {
+      "early_stopping": true,
+      "length_penalty": 2.0,
+      "max_length": 200,
+      "min_length": 30,
+      "no_repeat_ngram_size": 3,
+      "num_beams": 4,
+      "prefix": "summarize: "
+    },
+    "translation_en_to_de": {
+      "early_stopping": true,
+      "max_length": 300,
+      "num_beams": 4,
+      "prefix": "translate English to German: "
+    },
+    "translation_en_to_fr": {
+      "early_stopping": true,
+      "max_length": 300,
+      "num_beams": 4,
+      "prefix": "translate English to French: "
+    },
+    "translation_en_to_ro": {
+      "early_stopping": true,
+      "max_length": 300,
+      "num_beams": 4,
+      "prefix": "translate English to Romanian: "
+    }
+  },
+  "tie_word_embeddings": false,
+  "torch_dtype": "float32",
+  "transformers_version": "4.33.1",
+  "use_cache": true,
+  "vocab_size": 32128
+}