Upload folder using huggingface_hub

115_128_e3_3e-5/adapter_config.json

@@ -0,0 +1,39 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": null,
+  "base_model_name_or_path": "ibm-granite/granite-3.3-8b-base",
+  "bias": "none",
+  "corda_config": null,
+  "eva_config": null,
+  "exclude_modules": null,
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 256,
+  "lora_bias": false,
+  "lora_dropout": 0.05,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 128,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "gate_proj",
+    "down_proj",
+    "o_proj",
+    "up_proj",
+    "q_proj",
+    "v_proj",
+    "k_proj"
+  ],
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_dora": false,
+  "use_rslora": false
+}

115_128_e3_3e-5/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c00134c5388083c7506d12ca578d0772fda01286808aca2132445a8813a93ee9
+size 791751704

115_128_e3_3e-5/latest

@@ -0,0 +1 @@
+global_step825

115_128_e3_3e-5/rng_state_0.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d8ef25ccc7995222885a5125f3d1d431dbd4c7ab38e7a5301222d5f1ba60f622
+size 16389

115_128_e3_3e-5/rng_state_1.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:369616f975aa72725e66038adce44195885840bcd663c2e21fd50c331c9da787
+size 16389

115_128_e3_3e-5/rng_state_2.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6df6cb6d50ab93687077b1c82944957337afcc521c579258a468371ba1b61f97
+size 16389

115_128_e3_3e-5/rng_state_3.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a36eac1e5795a972273788fffb4b0febb1fcd709b5904a06dc4c13b7c86bc90d
+size 16389

115_128_e3_3e-5/rng_state_4.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1d19da65262f48df524ff8abbc108a8710297145d1d209bf6c9aa921b65a6e03
+size 16389

115_128_e3_3e-5/rng_state_5.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4004ce86656c819f1a89fa0f4c8de2c388738b521eed30b0703fe5082d509f0b
+size 16389

115_128_e3_3e-5/rng_state_6.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d688749bf4406f7ed0bb8e477f21833da470606278f9fc55c35360729faafca8
+size 16389

115_128_e3_3e-5/rng_state_7.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6bfd3000a92883cbf99cc6c35dd7cdf4e440fdf9a6bffcff8277904d56fb530f
+size 16389

115_128_e3_3e-5/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6e85cd94644bd1f4fd422c6833d47893de4994623f9cfbaa5efe5d23e0695a60
+size 1401

115_128_e3_3e-5/special_tokens_map.json

@@ -0,0 +1,45 @@
+{
+  "additional_special_tokens": [
+    "<|endoftext|>",
+    "<fim_prefix>",
+    "<fim_middle>",
+    "<fim_suffix>",
+    "<fim_pad>",
+    "<filename>",
+    "<gh_stars>",
+    "<issue_start>",
+    "<issue_comment>",
+    "<issue_closed>",
+    "<jupyter_start>",
+    "<jupyter_text>",
+    "<jupyter_code>",
+    "<jupyter_output>",
+    "<empty_output>",
+    "<commit_before>",
+    "<commit_msg>",
+    "<commit_after>",
+    "<reponame>"
+  ],
+  "bos_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": "<reponame>",
+  "unk_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

115_128_e3_3e-5/tokenizer_config.json

@@ -0,0 +1,188 @@
+{
+  "add_prefix_space": false,
+  "added_tokens_decoder": {
+    "0": {
+      "content": "<|endoftext|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "<fim_prefix>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "<fim_middle>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "3": {
+      "content": "<fim_suffix>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "4": {
+      "content": "<fim_pad>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "5": {
+      "content": "<filename>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "6": {
+      "content": "<gh_stars>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "7": {
+      "content": "<issue_start>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "8": {
+      "content": "<issue_comment>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "9": {
+      "content": "<issue_closed>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "10": {
+      "content": "<jupyter_start>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "11": {
+      "content": "<jupyter_text>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "12": {
+      "content": "<jupyter_code>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "13": {
+      "content": "<jupyter_output>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "14": {
+      "content": "<empty_output>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "15": {
+      "content": "<commit_before>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "16": {
+      "content": "<commit_msg>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "17": {
+      "content": "<commit_after>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "18": {
+      "content": "<reponame>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "additional_special_tokens": [
+    "<|endoftext|>",
+    "<fim_prefix>",
+    "<fim_middle>",
+    "<fim_suffix>",
+    "<fim_pad>",
+    "<filename>",
+    "<gh_stars>",
+    "<issue_start>",
+    "<issue_comment>",
+    "<issue_closed>",
+    "<jupyter_start>",
+    "<jupyter_text>",
+    "<jupyter_code>",
+    "<jupyter_output>",
+    "<empty_output>",
+    "<commit_before>",
+    "<commit_msg>",
+    "<commit_after>",
+    "<reponame>"
+  ],
+  "bos_token": "<|endoftext|>",
+  "clean_up_tokenization_spaces": true,
+  "eos_token": "<|endoftext|>",
+  "extra_special_tokens": {},
+  "model_max_length": 8192,
+  "pad_token": "<reponame>",
+  "padding_side": "left",
+  "tokenizer_class": "GPT2Tokenizer",
+  "unk_token": "<|endoftext|>",
+  "vocab_size": 49152
+}

115_128_e3_3e-5/trainer_state.json

@@ -0,0 +1,1189 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 825,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.018214936247723135,
+      "grad_norm": 1.1263623237609863,
+      "learning_rate": 2.8571428571428573e-06,
+      "loss": 1.3395,
+      "step": 5
+    },
+    {
+      "epoch": 0.03642987249544627,
+      "grad_norm": 0.7588332295417786,
+      "learning_rate": 6.428571428571429e-06,
+      "loss": 1.3413,
+      "step": 10
+    },
+    {
+      "epoch": 0.0546448087431694,
+      "grad_norm": 0.6433761715888977,
+      "learning_rate": 9.999999999999999e-06,
+      "loss": 1.3636,
+      "step": 15
+    },
+    {
+      "epoch": 0.07285974499089254,
+      "grad_norm": 0.5927166938781738,
+      "learning_rate": 1.3571428571428572e-05,
+      "loss": 1.3166,
+      "step": 20
+    },
+    {
+      "epoch": 0.09107468123861566,
+      "grad_norm": 0.7119314670562744,
+      "learning_rate": 1.7142857142857142e-05,
+      "loss": 1.2624,
+      "step": 25
+    },
+    {
+      "epoch": 0.1092896174863388,
+      "grad_norm": 0.5916720628738403,
+      "learning_rate": 2.0714285714285715e-05,
+      "loss": 1.2883,
+      "step": 30
+    },
+    {
+      "epoch": 0.12750455373406194,
+      "grad_norm": 0.5931192636489868,
+      "learning_rate": 2.4285714285714288e-05,
+      "loss": 1.2515,
+      "step": 35
+    },
+    {
+      "epoch": 0.14571948998178508,
+      "grad_norm": 0.6743040680885315,
+      "learning_rate": 2.7857142857142858e-05,
+      "loss": 1.1649,
+      "step": 40
+    },
+    {
+      "epoch": 0.16393442622950818,
+      "grad_norm": 0.7991889119148254,
+      "learning_rate": 2.9999517057811818e-05,
+      "loss": 1.1514,
+      "step": 45
+    },
+    {
+      "epoch": 0.18214936247723132,
+      "grad_norm": 0.6058378219604492,
+      "learning_rate": 2.9994084315326694e-05,
+      "loss": 1.1593,
+      "step": 50
+    },
+    {
+      "epoch": 0.20036429872495445,
+      "grad_norm": 0.6810833215713501,
+      "learning_rate": 2.9982617346265313e-05,
+      "loss": 1.0946,
+      "step": 55
+    },
+    {
+      "epoch": 0.2185792349726776,
+      "grad_norm": 0.7552086710929871,
+      "learning_rate": 2.9965120765400272e-05,
+      "loss": 1.0929,
+      "step": 60
+    },
+    {
+      "epoch": 0.23679417122040072,
+      "grad_norm": 0.6997225284576416,
+      "learning_rate": 2.9941601614064068e-05,
+      "loss": 1.0675,
+      "step": 65
+    },
+    {
+      "epoch": 0.2550091074681239,
+      "grad_norm": 0.7775698900222778,
+      "learning_rate": 2.9912069357315394e-05,
+      "loss": 1.0455,
+      "step": 70
+    },
+    {
+      "epoch": 0.273224043715847,
+      "grad_norm": 0.7171245217323303,
+      "learning_rate": 2.9876535880130008e-05,
+      "loss": 0.9769,
+      "step": 75
+    },
+    {
+      "epoch": 0.29143897996357016,
+      "grad_norm": 0.7159048318862915,
+      "learning_rate": 2.9835015482617726e-05,
+      "loss": 0.9351,
+      "step": 80
+    },
+    {
+      "epoch": 0.30965391621129323,
+      "grad_norm": 0.8580832481384277,
+      "learning_rate": 2.97875248742675e-05,
+      "loss": 0.8918,
+      "step": 85
+    },
+    {
+      "epoch": 0.32786885245901637,
+      "grad_norm": 0.8113096952438354,
+      "learning_rate": 2.9734083167222817e-05,
+      "loss": 0.8831,
+      "step": 90
+    },
+    {
+      "epoch": 0.3460837887067395,
+      "grad_norm": 0.9083524942398071,
+      "learning_rate": 2.9674711868590233e-05,
+      "loss": 0.8611,
+      "step": 95
+    },
+    {
+      "epoch": 0.36429872495446264,
+      "grad_norm": 1.0723042488098145,
+      "learning_rate": 2.9609434871784003e-05,
+      "loss": 0.8362,
+      "step": 100
+    },
+    {
+      "epoch": 0.3825136612021858,
+      "grad_norm": 0.8389668464660645,
+      "learning_rate": 2.9538278446910447e-05,
+      "loss": 0.7976,
+      "step": 105
+    },
+    {
+      "epoch": 0.4007285974499089,
+      "grad_norm": 0.9779384136199951,
+      "learning_rate": 2.946127123019575e-05,
+      "loss": 0.8156,
+      "step": 110
+    },
+    {
+      "epoch": 0.41894353369763204,
+      "grad_norm": 1.0124034881591797,
+      "learning_rate": 2.937844421246162e-05,
+      "loss": 0.7971,
+      "step": 115
+    },
+    {
+      "epoch": 0.4371584699453552,
+      "grad_norm": 0.9274071455001831,
+      "learning_rate": 2.9289830726653317e-05,
+      "loss": 0.7674,
+      "step": 120
+    },
+    {
+      "epoch": 0.4553734061930783,
+      "grad_norm": 1.1658202409744263,
+      "learning_rate": 2.919546643442511e-05,
+      "loss": 0.7071,
+      "step": 125
+    },
+    {
+      "epoch": 0.47358834244080145,
+      "grad_norm": 1.2840852737426758,
+      "learning_rate": 2.9095389311788626e-05,
+      "loss": 0.7052,
+      "step": 130
+    },
+    {
+      "epoch": 0.4918032786885246,
+      "grad_norm": 0.9791309833526611,
+      "learning_rate": 2.89896396338297e-05,
+      "loss": 0.7353,
+      "step": 135
+    },
+    {
+      "epoch": 0.5100182149362478,
+      "grad_norm": 1.0463306903839111,
+      "learning_rate": 2.8878259958500114e-05,
+      "loss": 0.7578,
+      "step": 140
+    },
+    {
+      "epoch": 0.5282331511839709,
+      "grad_norm": 1.0644102096557617,
+      "learning_rate": 2.8761295109490516e-05,
+      "loss": 0.6822,
+      "step": 145
+    },
+    {
+      "epoch": 0.546448087431694,
+      "grad_norm": 1.0078115463256836,
+      "learning_rate": 2.863879215819156e-05,
+      "loss": 0.6575,
+      "step": 150
+    },
+    {
+      "epoch": 0.5646630236794171,
+      "grad_norm": 1.0707165002822876,
+      "learning_rate": 2.851080040475049e-05,
+      "loss": 0.6041,
+      "step": 155
+    },
+    {
+      "epoch": 0.5828779599271403,
+      "grad_norm": 1.088377833366394,
+      "learning_rate": 2.8377371358230733e-05,
+      "loss": 0.5583,
+      "step": 160
+    },
+    {
+      "epoch": 0.6010928961748634,
+      "grad_norm": 1.2920998334884644,
+      "learning_rate": 2.82385587158826e-05,
+      "loss": 0.6297,
+      "step": 165
+    },
+    {
+      "epoch": 0.6193078324225865,
+      "grad_norm": 1.2502070665359497,
+      "learning_rate": 2.8094418341533313e-05,
+      "loss": 0.5495,
+      "step": 170
+    },
+    {
+      "epoch": 0.6375227686703097,
+      "grad_norm": 1.1710216999053955,
+      "learning_rate": 2.7945008243105147e-05,
+      "loss": 0.5821,
+      "step": 175
+    },
+    {
+      "epoch": 0.6557377049180327,
+      "grad_norm": 1.3169817924499512,
+      "learning_rate": 2.7790388549270703e-05,
+      "loss": 0.5166,
+      "step": 180
+    },
+    {
+      "epoch": 0.6739526411657559,
+      "grad_norm": 1.1689430475234985,
+      "learning_rate": 2.763062148525469e-05,
+      "loss": 0.5492,
+      "step": 185
+    },
+    {
+      "epoch": 0.692167577413479,
+      "grad_norm": 1.170713186264038,
+      "learning_rate": 2.7465771347791953e-05,
+      "loss": 0.5104,
+      "step": 190
+    },
+    {
+      "epoch": 0.7103825136612022,
+      "grad_norm": 1.1507899761199951,
+      "learning_rate": 2.7295904479251917e-05,
+      "loss": 0.4871,
+      "step": 195
+    },
+    {
+      "epoch": 0.7285974499089253,
+      "grad_norm": 1.2234132289886475,
+      "learning_rate": 2.7121089240939674e-05,
+      "loss": 0.4879,
+      "step": 200
+    },
+    {
+      "epoch": 0.7468123861566485,
+      "grad_norm": 1.7721701860427856,
+      "learning_rate": 2.6941395985584656e-05,
+      "loss": 0.4308,
+      "step": 205
+    },
+    {
+      "epoch": 0.7650273224043715,
+      "grad_norm": 1.1693341732025146,
+      "learning_rate": 2.6756897029027853e-05,
+      "loss": 0.4932,
+      "step": 210
+    },
+    {
+      "epoch": 0.7832422586520947,
+      "grad_norm": 2.074277639389038,
+      "learning_rate": 2.656766662111897e-05,
+      "loss": 0.4544,
+      "step": 215
+    },
+    {
+      "epoch": 0.8014571948998178,
+      "grad_norm": 1.084118366241455,
+      "learning_rate": 2.6373780915835296e-05,
+      "loss": 0.4354,
+      "step": 220
+    },
+    {
+      "epoch": 0.819672131147541,
+      "grad_norm": 1.283940076828003,
+      "learning_rate": 2.617531794063427e-05,
+      "loss": 0.4369,
+      "step": 225
+    },
+    {
+      "epoch": 0.8378870673952641,
+      "grad_norm": 1.4470164775848389,
+      "learning_rate": 2.597235756505203e-05,
+      "loss": 0.4074,
+      "step": 230
+    },
+    {
+      "epoch": 0.8561020036429873,
+      "grad_norm": 1.157853603363037,
+      "learning_rate": 2.5764981468560743e-05,
+      "loss": 0.4677,
+      "step": 235
+    },
+    {
+      "epoch": 0.8743169398907104,
+      "grad_norm": 1.1284898519515991,
+      "learning_rate": 2.555327310769742e-05,
+      "loss": 0.3683,
+      "step": 240
+    },
+    {
+      "epoch": 0.8925318761384335,
+      "grad_norm": 1.2154994010925293,
+      "learning_rate": 2.5337317682477666e-05,
+      "loss": 0.4181,
+      "step": 245
+    },
+    {
+      "epoch": 0.9107468123861566,
+      "grad_norm": 1.3954899311065674,
+      "learning_rate": 2.5117202102107707e-05,
+      "loss": 0.3811,
+      "step": 250
+    },
+    {
+      "epoch": 0.9289617486338798,
+      "grad_norm": 1.1357766389846802,
+      "learning_rate": 2.4893014950008662e-05,
+      "loss": 0.3554,
+      "step": 255
+    },
+    {
+      "epoch": 0.9471766848816029,
+      "grad_norm": 1.2787904739379883,
+      "learning_rate": 2.4664846448166925e-05,
+      "loss": 0.3589,
+      "step": 260
+    },
+    {
+      "epoch": 0.9653916211293261,
+      "grad_norm": 1.125148892402649,
+      "learning_rate": 2.4432788420825228e-05,
+      "loss": 0.3407,
+      "step": 265
+    },
+    {
+      "epoch": 0.9836065573770492,
+      "grad_norm": 1.4910621643066406,
+      "learning_rate": 2.4196934257528826e-05,
+      "loss": 0.3628,
+      "step": 270
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9073554277420044,
+      "learning_rate": 2.3957378875541795e-05,
+      "loss": 0.3028,
+      "step": 275
+    },
+    {
+      "epoch": 1.018214936247723,
+      "grad_norm": 1.1472718715667725,
+      "learning_rate": 2.3714218681648423e-05,
+      "loss": 0.2949,
+      "step": 280
+    },
+    {
+      "epoch": 1.0364298724954462,
+      "grad_norm": 1.2372900247573853,
+      "learning_rate": 2.3467551533355283e-05,
+      "loss": 0.282,
+      "step": 285
+    },
+    {
+      "epoch": 1.0546448087431695,
+      "grad_norm": 1.1617836952209473,
+      "learning_rate": 2.3217476699509373e-05,
+      "loss": 0.2824,
+      "step": 290
+    },
+    {
+      "epoch": 1.0728597449908925,
+      "grad_norm": 1.2221508026123047,
+      "learning_rate": 2.2964094820348302e-05,
+      "loss": 0.2804,
+      "step": 295
+    },
+    {
+      "epoch": 1.0910746812386156,
+      "grad_norm": 1.3465958833694458,
+      "learning_rate": 2.2707507866998633e-05,
+      "loss": 0.2817,
+      "step": 300
+    },
+    {
+      "epoch": 1.1092896174863387,
+      "grad_norm": 1.2395645380020142,
+      "learning_rate": 2.2447819100438518e-05,
+      "loss": 0.2656,
+      "step": 305
+    },
+    {
+      "epoch": 1.127504553734062,
+      "grad_norm": 1.0963815450668335,
+      "learning_rate": 2.2185133029941333e-05,
+      "loss": 0.2463,
+      "step": 310
+    },
+    {
+      "epoch": 1.145719489981785,
+      "grad_norm": 1.3694543838500977,
+      "learning_rate": 2.1919555371016914e-05,
+      "loss": 0.2723,
+      "step": 315
+    },
+    {
+      "epoch": 1.1639344262295082,
+      "grad_norm": 1.1197879314422607,
+      "learning_rate": 2.165119300286733e-05,
+      "loss": 0.2338,
+      "step": 320
+    },
+    {
+      "epoch": 1.1821493624772312,
+      "grad_norm": 1.1068480014801025,
+      "learning_rate": 2.13801539253744e-05,
+      "loss": 0.2832,
+      "step": 325
+    },
+    {
+      "epoch": 1.2003642987249545,
+      "grad_norm": 1.1547338962554932,
+      "learning_rate": 2.110654721563615e-05,
+      "loss": 0.2588,
+      "step": 330
+    },
+    {
+      "epoch": 1.2185792349726776,
+      "grad_norm": 1.1785893440246582,
+      "learning_rate": 2.0830482984069763e-05,
+      "loss": 0.2521,
+      "step": 335
+    },
+    {
+      "epoch": 1.2367941712204007,
+      "grad_norm": 1.1551320552825928,
+      "learning_rate": 2.055207233009872e-05,
+      "loss": 0.1987,
+      "step": 340
+    },
+    {
+      "epoch": 1.255009107468124,
+      "grad_norm": 1.0682295560836792,
+      "learning_rate": 2.027142729744184e-05,
+      "loss": 0.2354,
+      "step": 345
+    },
+    {
+      "epoch": 1.273224043715847,
+      "grad_norm": 1.2101492881774902,
+      "learning_rate": 1.998866082902243e-05,
+      "loss": 0.2638,
+      "step": 350
+    },
+    {
+      "epoch": 1.2914389799635702,
+      "grad_norm": 1.1589305400848389,
+      "learning_rate": 1.9703886721515405e-05,
+      "loss": 0.253,
+      "step": 355
+    },
+    {
+      "epoch": 1.3096539162112932,
+      "grad_norm": 1.2090011835098267,
+      "learning_rate": 1.9417219579550937e-05,
+      "loss": 0.2302,
+      "step": 360
+    },
+    {
+      "epoch": 1.3278688524590163,
+      "grad_norm": 1.1908096075057983,
+      "learning_rate": 1.9128774769592904e-05,
+      "loss": 0.1855,
+      "step": 365
+    },
+    {
+      "epoch": 1.3460837887067396,
+      "grad_norm": 1.027683973312378,
+      "learning_rate": 1.8838668373510768e-05,
+      "loss": 0.1979,
+      "step": 370
+    },
+    {
+      "epoch": 1.3642987249544627,
+      "grad_norm": 1.1715365648269653,
+      "learning_rate": 1.8547017141863497e-05,
+      "loss": 0.2113,
+      "step": 375
+    },
+    {
+      "epoch": 1.3825136612021858,
+      "grad_norm": 1.154416561126709,
+      "learning_rate": 1.825393844691445e-05,
+      "loss": 0.1986,
+      "step": 380
+    },
+    {
+      "epoch": 1.4007285974499089,
+      "grad_norm": 1.1075443029403687,
+      "learning_rate": 1.7959550235396002e-05,
+      "loss": 0.1922,
+      "step": 385
+    },
+    {
+      "epoch": 1.418943533697632,
+      "grad_norm": 1.0319886207580566,
+      "learning_rate": 1.766397098104302e-05,
+      "loss": 0.1938,
+      "step": 390
+    },
+    {
+      "epoch": 1.4371584699453552,
+      "grad_norm": 1.608643651008606,
+      "learning_rate": 1.7367319636914218e-05,
+      "loss": 0.1955,
+      "step": 395
+    },
+    {
+      "epoch": 1.4553734061930783,
+      "grad_norm": 1.1023911237716675,
+      "learning_rate": 1.7069715587520648e-05,
+      "loss": 0.155,
+      "step": 400
+    },
+    {
+      "epoch": 1.4735883424408014,
+      "grad_norm": 1.079603672027588,
+      "learning_rate": 1.6771278600780533e-05,
+      "loss": 0.172,
+      "step": 405
+    },
+    {
+      "epoch": 1.4918032786885247,
+      "grad_norm": 1.3646568059921265,
+      "learning_rate": 1.64721287798198e-05,
+      "loss": 0.1709,
+      "step": 410
+    },
+    {
+      "epoch": 1.5100182149362478,
+      "grad_norm": 0.9472083449363708,
+      "learning_rate": 1.6172386514637726e-05,
+      "loss": 0.1541,
+      "step": 415
+    },
+    {
+      "epoch": 1.5282331511839709,
+      "grad_norm": 1.1797008514404297,
+      "learning_rate": 1.587217243365714e-05,
+      "loss": 0.1809,
+      "step": 420
+    },
+    {
+      "epoch": 1.5464480874316942,
+      "grad_norm": 1.1122612953186035,
+      "learning_rate": 1.5571607355178632e-05,
+      "loss": 0.1416,
+      "step": 425
+    },
+    {
+      "epoch": 1.564663023679417,
+      "grad_norm": 1.2478797435760498,
+      "learning_rate": 1.5270812238758407e-05,
+      "loss": 0.1722,
+      "step": 430
+    },
+    {
+      "epoch": 1.5828779599271403,
+      "grad_norm": 1.1090872287750244,
+      "learning_rate": 1.4969908136529254e-05,
+      "loss": 0.1754,
+      "step": 435
+    },
+    {
+      "epoch": 1.6010928961748634,
+      "grad_norm": 1.0214531421661377,
+      "learning_rate": 1.4669016144484254e-05,
+      "loss": 0.1397,
+      "step": 440
+    },
+    {
+      "epoch": 1.6193078324225865,
+      "grad_norm": 1.0101455450057983,
+      "learning_rate": 1.4368257353742863e-05,
+      "loss": 0.1432,
+      "step": 445
+    },
+    {
+      "epoch": 1.6375227686703098,
+      "grad_norm": 1.2055903673171997,
+      "learning_rate": 1.4067752801818948e-05,
+      "loss": 0.148,
+      "step": 450
+    },
+    {
+      "epoch": 1.6557377049180326,
+      "grad_norm": 1.2197571992874146,
+      "learning_rate": 1.3767623423910369e-05,
+      "loss": 0.1389,
+      "step": 455
+    },
+    {
+      "epoch": 1.673952641165756,
+      "grad_norm": 1.0878300666809082,
+      "learning_rate": 1.3467990004229784e-05,
+      "loss": 0.1496,
+      "step": 460
+    },
+    {
+      "epoch": 1.692167577413479,
+      "grad_norm": 1.0375535488128662,
+      "learning_rate": 1.316897312739618e-05,
+      "loss": 0.1331,
+      "step": 465
+    },
+    {
+      "epoch": 1.710382513661202,
+      "grad_norm": 1.1559418439865112,
+      "learning_rate": 1.2870693129906725e-05,
+      "loss": 0.1281,
+      "step": 470
+    },
+    {
+      "epoch": 1.7285974499089254,
+      "grad_norm": 1.1000542640686035,
+      "learning_rate": 1.2573270051708529e-05,
+      "loss": 0.1521,
+      "step": 475
+    },
+    {
+      "epoch": 1.7468123861566485,
+      "grad_norm": 1.2957032918930054,
+      "learning_rate": 1.2276823587889675e-05,
+      "loss": 0.1258,
+      "step": 480
+    },
+    {
+      "epoch": 1.7650273224043715,
+      "grad_norm": 1.0486924648284912,
+      "learning_rate": 1.1981473040509078e-05,
+      "loss": 0.1411,
+      "step": 485
+    },
+    {
+      "epoch": 1.7832422586520948,
+      "grad_norm": 1.1459728479385376,
+      "learning_rate": 1.1687337270584518e-05,
+      "loss": 0.1266,
+      "step": 490
+    },
+    {
+      "epoch": 1.8014571948998177,
+      "grad_norm": 1.2694897651672363,
+      "learning_rate": 1.139453465025814e-05,
+      "loss": 0.1317,
+      "step": 495
+    },
+    {
+      "epoch": 1.819672131147541,
+      "grad_norm": 1.0525625944137573,
+      "learning_rate": 1.1103183015158704e-05,
+      "loss": 0.1428,
+      "step": 500
+    },
+    {
+      "epoch": 1.837887067395264,
+      "grad_norm": 0.9716005921363831,
+      "learning_rate": 1.08133996169798e-05,
+      "loss": 0.1102,
+      "step": 505
+    },
+    {
+      "epoch": 1.8561020036429872,
+      "grad_norm": 1.0540605783462524,
+      "learning_rate": 1.0525301076292959e-05,
+      "loss": 0.115,
+      "step": 510
+    },
+    {
+      "epoch": 1.8743169398907105,
+      "grad_norm": 0.9393624663352966,
+      "learning_rate": 1.0239003335614867e-05,
+      "loss": 0.1137,
+      "step": 515
+    },
+    {
+      "epoch": 1.8925318761384335,
+      "grad_norm": 0.9130032062530518,
+      "learning_rate": 9.954621612747371e-06,
+      "loss": 0.1217,
+      "step": 520
+    },
+    {
+      "epoch": 1.9107468123861566,
+      "grad_norm": 1.033608317375183,
+      "learning_rate": 9.67227035440918e-06,
+      "loss": 0.1156,
+      "step": 525
+    },
+    {
+      "epoch": 1.92896174863388,
+      "grad_norm": 0.8388630747795105,
+      "learning_rate": 9.392063190177873e-06,
+      "loss": 0.1064,
+      "step": 530
+    },
+    {
+      "epoch": 1.9471766848816028,
+      "grad_norm": 0.9510135650634766,
+      "learning_rate": 9.114112886760748e-06,
+      "loss": 0.106,
+      "step": 535
+    },
+    {
+      "epoch": 1.965391621129326,
+      "grad_norm": 0.9661350250244141,
+      "learning_rate": 8.838531302612942e-06,
+      "loss": 0.1122,
+      "step": 540
+    },
+    {
+      "epoch": 1.9836065573770492,
+      "grad_norm": 1.0655405521392822,
+      "learning_rate": 8.565429342921069e-06,
+      "loss": 0.1167,
+      "step": 545
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.5059938430786133,
+      "learning_rate": 8.294916914970466e-06,
+      "loss": 0.1006,
+      "step": 550
+    },
+    {
+      "epoch": 2.0182149362477233,
+      "grad_norm": 0.8874540328979492,
+      "learning_rate": 8.027102883914053e-06,
+      "loss": 0.0826,
+      "step": 555
+    },
+    {
+      "epoch": 2.036429872495446,
+      "grad_norm": 0.7909408211708069,
+      "learning_rate": 7.762095028960601e-06,
+      "loss": 0.0811,
+      "step": 560
+    },
+    {
+      "epoch": 2.0546448087431695,
+      "grad_norm": 0.7704826593399048,
+      "learning_rate": 7.500000000000004e-06,
+      "loss": 0.0802,
+      "step": 565
+    },
+    {
+      "epoch": 2.0728597449908923,
+      "grad_norm": 0.7435580492019653,
+      "learning_rate": 7.2409232746830405e-06,
+      "loss": 0.0793,
+      "step": 570
+    },
+    {
+      "epoch": 2.0910746812386156,
+      "grad_norm": 0.7861345410346985,
+      "learning_rate": 6.984969115972913e-06,
+      "loss": 0.0764,
+      "step": 575
+    },
+    {
+      "epoch": 2.109289617486339,
+      "grad_norm": 0.9503616690635681,
+      "learning_rate": 6.732240530185607e-06,
+      "loss": 0.085,
+      "step": 580
+    },
+    {
+      "epoch": 2.1275045537340618,
+      "grad_norm": 0.6551730036735535,
+      "learning_rate": 6.482839225535962e-06,
+      "loss": 0.0775,
+      "step": 585
+    },
+    {
+      "epoch": 2.145719489981785,
+      "grad_norm": 0.9471770524978638,
+      "learning_rate": 6.2368655712061866e-06,
+      "loss": 0.0746,
+      "step": 590
+    },
+    {
+      "epoch": 2.1639344262295084,
+      "grad_norm": 0.8260408639907837,
+      "learning_rate": 5.994418556953226e-06,
+      "loss": 0.0734,
+      "step": 595
+    },
+    {
+      "epoch": 2.1821493624772312,
+      "grad_norm": 0.8403226137161255,
+      "learning_rate": 5.755595753271282e-06,
+      "loss": 0.0825,
+      "step": 600
+    },
+    {
+      "epoch": 2.2003642987249545,
+      "grad_norm": 0.8162021636962891,
+      "learning_rate": 5.520493272125489e-06,
+      "loss": 0.0771,
+      "step": 605
+    },
+    {
+      "epoch": 2.2185792349726774,
+      "grad_norm": 0.8946626782417297,
+      "learning_rate": 5.289205728272587e-06,
+      "loss": 0.0806,
+      "step": 610
+    },
+    {
+      "epoch": 2.2367941712204007,
+      "grad_norm": 0.8603159189224243,
+      "learning_rate": 5.061826201184078e-06,
+      "loss": 0.0776,
+      "step": 615
+    },
+    {
+      "epoch": 2.255009107468124,
+      "grad_norm": 0.7660353779792786,
+      "learning_rate": 4.838446197587313e-06,
+      "loss": 0.0774,
+      "step": 620
+    },
+    {
+      "epoch": 2.273224043715847,
+      "grad_norm": 0.7102609872817993,
+      "learning_rate": 4.619155614639457e-06,
+      "loss": 0.0713,
+      "step": 625
+    },
+    {
+      "epoch": 2.29143897996357,
+      "grad_norm": 0.7758877277374268,
+      "learning_rate": 4.404042703749258e-06,
+      "loss": 0.0697,
+      "step": 630
+    },
+    {
+      "epoch": 2.3096539162112935,
+      "grad_norm": 0.87920081615448,
+      "learning_rate": 4.19319403506107e-06,
+      "loss": 0.0728,
+      "step": 635
+    },
+    {
+      "epoch": 2.3278688524590163,
+      "grad_norm": 0.7006933689117432,
+      "learning_rate": 3.986694462615552e-06,
+      "loss": 0.0802,
+      "step": 640
+    },
+    {
+      "epoch": 2.3460837887067396,
+      "grad_norm": 0.6575765609741211,
+      "learning_rate": 3.7846270902009695e-06,
+      "loss": 0.0673,
+      "step": 645
+    },
+    {
+      "epoch": 2.3642987249544625,
+      "grad_norm": 0.852051854133606,
+      "learning_rate": 3.5870732379088407e-06,
+      "loss": 0.0744,
+      "step": 650
+    },
+    {
+      "epoch": 2.3825136612021858,
+      "grad_norm": 0.7440199255943298,
+      "learning_rate": 3.394112409407455e-06,
+      "loss": 0.0712,
+      "step": 655
+    },
+    {
+      "epoch": 2.400728597449909,
+      "grad_norm": 1.0174368619918823,
+      "learning_rate": 3.2058222599463863e-06,
+      "loss": 0.0761,
+      "step": 660
+    },
+    {
+      "epoch": 2.418943533697632,
+      "grad_norm": 0.8462310433387756,
+      "learning_rate": 3.0222785651048516e-06,
+      "loss": 0.0711,
+      "step": 665
+    },
+    {
+      "epoch": 2.4371584699453552,
+      "grad_norm": 0.6031699776649475,
+      "learning_rate": 2.843555190296568e-06,
+      "loss": 0.0645,
+      "step": 670
+    },
+    {
+      "epoch": 2.455373406193078,
+      "grad_norm": 0.6558037996292114,
+      "learning_rate": 2.66972406104333e-06,
+      "loss": 0.0651,
+      "step": 675
+    },
+    {
+      "epoch": 2.4735883424408014,
+      "grad_norm": 0.6522188782691956,
+      "learning_rate": 2.50085513402924e-06,
+      "loss": 0.0679,
+      "step": 680
+    },
+    {
+      "epoch": 2.4918032786885247,
+      "grad_norm": 0.6738578081130981,
+      "learning_rate": 2.3370163689473232e-06,
+      "loss": 0.062,
+      "step": 685
+    },
+    {
+      "epoch": 2.510018214936248,
+      "grad_norm": 0.7459775805473328,
+      "learning_rate": 2.1782737011497992e-06,
+      "loss": 0.0677,
+      "step": 690
+    },
+    {
+      "epoch": 2.528233151183971,
+      "grad_norm": 0.7526170611381531,
+      "learning_rate": 2.0246910151130204e-06,
+      "loss": 0.0731,
+      "step": 695
+    },
+    {
+      "epoch": 2.546448087431694,
+      "grad_norm": 0.652023434638977,
+      "learning_rate": 1.8763301187277554e-06,
+      "loss": 0.0583,
+      "step": 700
+    },
+    {
+      "epoch": 2.564663023679417,
+      "grad_norm": 0.6729693412780762,
+      "learning_rate": 1.7332507184252317e-06,
+      "loss": 0.0675,
+      "step": 705
+    },
+    {
+      "epoch": 2.5828779599271403,
+      "grad_norm": 0.8072725534439087,
+      "learning_rate": 1.5955103951488177e-06,
+      "loss": 0.0746,
+      "step": 710
+    },
+    {
+      "epoch": 2.6010928961748636,
+      "grad_norm": 0.5885034203529358,
+      "learning_rate": 1.4631645811811257e-06,
+      "loss": 0.0613,
+      "step": 715
+    },
+    {
+      "epoch": 2.6193078324225865,
+      "grad_norm": 0.7038059234619141,
+      "learning_rate": 1.3362665378358253e-06,
+      "loss": 0.0787,
+      "step": 720
+    },
+    {
+      "epoch": 2.6375227686703098,
+      "grad_norm": 0.5660616159439087,
+      "learning_rate": 1.2148673340231353e-06,
+      "loss": 0.0502,
+      "step": 725
+    },
+    {
+      "epoch": 2.6557377049180326,
+      "grad_norm": 0.622959315776825,
+      "learning_rate": 1.099015825697608e-06,
+      "loss": 0.0607,
+      "step": 730
+    },
+    {
+      "epoch": 2.673952641165756,
+      "grad_norm": 0.6994828581809998,
+      "learning_rate": 9.887586361965422e-07,
+      "loss": 0.0601,
+      "step": 735
+    },
+    {
+      "epoch": 2.6921675774134792,
+      "grad_norm": 0.4787428677082062,
+      "learning_rate": 8.841401374768437e-07,
+      "loss": 0.0611,
+      "step": 740
+    },
+    {
+      "epoch": 2.710382513661202,
+      "grad_norm": 0.7188135385513306,
+      "learning_rate": 7.852024322579649e-07,
+      "loss": 0.0648,
+      "step": 745
+    },
+    {
+      "epoch": 2.7285974499089254,
+      "grad_norm": 0.550809919834137,
+      "learning_rate": 6.919853370780554e-07,
+      "loss": 0.0555,
+      "step": 750
+    },
+    {
+      "epoch": 2.7468123861566482,
+      "grad_norm": 0.576739490032196,
+      "learning_rate": 6.045263662701905e-07,
+      "loss": 0.0592,
+      "step": 755
+    },
+    {
+      "epoch": 2.7650273224043715,
+      "grad_norm": 0.7706507444381714,
+      "learning_rate": 5.228607168650701e-07,
+      "loss": 0.0586,
+      "step": 760
+    },
+    {
+      "epoch": 2.783242258652095,
+      "grad_norm": 0.6334231495857239,
+      "learning_rate": 4.4702125442631304e-07,
+      "loss": 0.0575,
+      "step": 765
+    },
+    {
+      "epoch": 2.8014571948998177,
+      "grad_norm": 0.6402108669281006,
+      "learning_rate": 3.770384998240306e-07,
+      "loss": 0.0666,
+      "step": 770
+    },
+    {
+      "epoch": 2.819672131147541,
+      "grad_norm": 0.6130778193473816,
+      "learning_rate": 3.129406169519816e-07,
+      "loss": 0.0607,
+      "step": 775
+    },
+    {
+      "epoch": 2.837887067395264,
+      "grad_norm": 0.6811858415603638,
+      "learning_rate": 2.5475340139328897e-07,
+      "loss": 0.072,
+      "step": 780
+    },
+    {
+      "epoch": 2.856102003642987,
+      "grad_norm": 0.6625553965568542,
+      "learning_rate": 2.0250027003925364e-07,
+      "loss": 0.064,
+      "step": 785
+    },
+    {
+      "epoch": 2.8743169398907105,
+      "grad_norm": 0.5846689939498901,
+      "learning_rate": 1.5620225166544155e-07,
+      "loss": 0.0626,
+      "step": 790
+    },
+    {
+      "epoch": 2.8925318761384338,
+      "grad_norm": 0.7103590965270996,
+      "learning_rate": 1.1587797846885628e-07,
+      "loss": 0.0583,
+      "step": 795
+    },
+    {
+      "epoch": 2.9107468123861566,
+      "grad_norm": 0.7647429704666138,
+      "learning_rate": 8.154367856959887e-08,
+      "loss": 0.0615,
+      "step": 800
+    },
+    {
+      "epoch": 2.92896174863388,
+      "grad_norm": 0.6568042635917664,
+      "learning_rate": 5.3213169479996994e-08,
+      "loss": 0.0696,
+      "step": 805
+    },
+    {
+      "epoch": 2.947176684881603,
+      "grad_norm": 0.6124881505966187,
+      "learning_rate": 3.0897852543892993e-08,
+      "loss": 0.061,
+      "step": 810
+    },
+    {
+      "epoch": 2.965391621129326,
+      "grad_norm": 0.6192994117736816,
+      "learning_rate": 1.4606708348275932e-08,
+      "loss": 0.0598,
+      "step": 815
+    },
+    {
+      "epoch": 2.9836065573770494,
+      "grad_norm": 0.5998212695121765,
+      "learning_rate": 4.346293109130928e-09,
+      "loss": 0.0635,
+      "step": 820
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.666495680809021,
+      "learning_rate": 1.2073603295303937e-10,
+      "loss": 0.0662,
+      "step": 825
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 825,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 2000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.0672591098850836e+18,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}

115_128_e3_3e-5/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6a8642741574d20a570268407c1351f1674824f698a33b2cdc26c9292bd33db4
+size 8209

115_128_e3_3e-5/zero_to_fp32.py

@@ -0,0 +1,604 @@
+#!/usr/bin/env python
+
+# Copyright (c) Microsoft Corporation.
+# SPDX-License-Identifier: Apache-2.0
+
+# DeepSpeed Team
+
+# This script extracts fp32 consolidated weights from a zero 1, 2 and 3 DeepSpeed checkpoints. It gets
+# copied into the top level checkpoint dir, so the user can easily do the conversion at any point in
+# the future. Once extracted, the weights don't require DeepSpeed and can be used in any
+# application.
+#
+# example: python zero_to_fp32.py . pytorch_model.bin
+
+import argparse
+import torch
+import glob
+import math
+import os
+import re
+from collections import OrderedDict
+from dataclasses import dataclass
+
+# while this script doesn't use deepspeed to recover data, since the checkpoints are pickled with
+# DeepSpeed data structures it has to be available in the current python environment.
+from deepspeed.utils import logger
+from deepspeed.checkpoint.constants import (DS_VERSION, OPTIMIZER_STATE_DICT, SINGLE_PARTITION_OF_FP32_GROUPS,
+                                            FP32_FLAT_GROUPS, ZERO_STAGE, PARTITION_COUNT, PARAM_SHAPES, BUFFER_NAMES,
+                                            FROZEN_PARAM_SHAPES, FROZEN_PARAM_FRAGMENTS)
+
+
+@dataclass
+class zero_model_state:
+    buffers: dict()
+    param_shapes: dict()
+    shared_params: list
+    ds_version: int
+    frozen_param_shapes: dict()
+    frozen_param_fragments: dict()
+
+
+debug = 0
+
+# load to cpu
+device = torch.device('cpu')
+
+
+def atoi(text):
+    return int(text) if text.isdigit() else text
+
+
+def natural_keys(text):
+    '''
+    alist.sort(key=natural_keys) sorts in human order
+    http://nedbatchelder.com/blog/200712/human_sorting.html
+    (See Toothy's implementation in the comments)
+    '''
+    return [atoi(c) for c in re.split(r'(\d+)', text)]
+
+
+def get_model_state_file(checkpoint_dir, zero_stage):
+    if not os.path.isdir(checkpoint_dir):
+        raise FileNotFoundError(f"Directory '{checkpoint_dir}' doesn't exist")
+
+    # there should be only one file
+    if zero_stage <= 2:
+        file = os.path.join(checkpoint_dir, "mp_rank_00_model_states.pt")
+    elif zero_stage == 3:
+        file = os.path.join(checkpoint_dir, "zero_pp_rank_0_mp_rank_00_model_states.pt")
+
+    if not os.path.exists(file):
+        raise FileNotFoundError(f"can't find model states file at '{file}'")
+
+    return file
+
+
+def get_checkpoint_files(checkpoint_dir, glob_pattern):
+    # XXX: need to test that this simple glob rule works for multi-node setup too
+    ckpt_files = sorted(glob.glob(os.path.join(checkpoint_dir, glob_pattern)), key=natural_keys)
+
+    if len(ckpt_files) == 0:
+        raise FileNotFoundError(f"can't find {glob_pattern} files in directory '{checkpoint_dir}'")
+
+    return ckpt_files
+
+
+def get_optim_files(checkpoint_dir):
+    return get_checkpoint_files(checkpoint_dir, "*_optim_states.pt")
+
+
+def get_model_state_files(checkpoint_dir):
+    return get_checkpoint_files(checkpoint_dir, "*_model_states.pt")
+
+
+def parse_model_states(files):
+    zero_model_states = []
+    for file in files:
+        state_dict = torch.load(file, map_location=device)
+
+        if BUFFER_NAMES not in state_dict:
+            raise ValueError(f"{file} is not a model state checkpoint")
+        buffer_names = state_dict[BUFFER_NAMES]
+        if debug:
+            print("Found buffers:", buffer_names)
+
+        # recover just the buffers while restoring them to fp32 if they were saved in fp16
+        buffers = {k: v.float() for k, v in state_dict["module"].items() if k in buffer_names}
+        param_shapes = state_dict[PARAM_SHAPES]
+
+        # collect parameters that are included in param_shapes
+        param_names = []
+        for s in param_shapes:
+            for name in s.keys():
+                param_names.append(name)
+
+        # update with frozen parameters
+        frozen_param_shapes = state_dict.get(FROZEN_PARAM_SHAPES, None)
+        if frozen_param_shapes is not None:
+            if debug:
+                print(f"Found frozen_param_shapes: {frozen_param_shapes}")
+            param_names += list(frozen_param_shapes.keys())
+
+        # handle shared params
+        shared_params = [[k, v] for k, v in state_dict["shared_params"].items()]
+
+        ds_version = state_dict.get(DS_VERSION, None)
+
+        frozen_param_fragments = state_dict.get(FROZEN_PARAM_FRAGMENTS, None)
+
+        z_model_state = zero_model_state(buffers=buffers,
+                                         param_shapes=param_shapes,
+                                         shared_params=shared_params,
+                                         ds_version=ds_version,
+                                         frozen_param_shapes=frozen_param_shapes,
+                                         frozen_param_fragments=frozen_param_fragments)
+        zero_model_states.append(z_model_state)
+
+    return zero_model_states
+
+
+def parse_optim_states(files, ds_checkpoint_dir):
+
+    total_files = len(files)
+    state_dicts = []
+    for f in files:
+        state_dict = torch.load(f, map_location=device)
+        # immediately discard the potentially huge 2 optimizer states as we only care for fp32 master weights
+        # and also handle the case where it was already removed by another helper script
+        state_dict["optimizer_state_dict"].pop("optimizer_state_dict", None)
+        state_dicts.append(state_dict)
+
+    if not ZERO_STAGE in state_dicts[0][OPTIMIZER_STATE_DICT]:
+        raise ValueError(f"{files[0]} is not a zero checkpoint")
+    zero_stage = state_dicts[0][OPTIMIZER_STATE_DICT][ZERO_STAGE]
+    world_size = state_dicts[0][OPTIMIZER_STATE_DICT][PARTITION_COUNT]
+
+    # For ZeRO-2 each param group can have different partition_count as data parallelism for expert
+    # parameters can be different from data parallelism for non-expert parameters. So we can just
+    # use the max of the partition_count to get the dp world_size.
+
+    if type(world_size) is list:
+        world_size = max(world_size)
+
+    if world_size != total_files:
+        raise ValueError(
+            f"Expected {world_size} of '*_optim_states.pt' under '{ds_checkpoint_dir}' but found {total_files} files. "
+            "Possibly due to an overwrite of an old checkpoint, or a checkpoint didn't get saved by one or more processes."
+        )
+
+    # the groups are named differently in each stage
+    if zero_stage <= 2:
+        fp32_groups_key = SINGLE_PARTITION_OF_FP32_GROUPS
+    elif zero_stage == 3:
+        fp32_groups_key = FP32_FLAT_GROUPS
+    else:
+        raise ValueError(f"unknown zero stage {zero_stage}")
+
+    if zero_stage <= 2:
+        fp32_flat_groups = [state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key] for i in range(len(state_dicts))]
+    elif zero_stage == 3:
+        # if there is more than one param group, there will be multiple flattened tensors - one
+        # flattened tensor per group - for simplicity merge them into a single tensor
+        #
+        # XXX: could make the script more memory efficient for when there are multiple groups - it
+        # will require matching the sub-lists of param_shapes for each param group flattened tensor
+
+        fp32_flat_groups = [
+            torch.cat(state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key], 0) for i in range(len(state_dicts))
+        ]
+
+    return zero_stage, world_size, fp32_flat_groups
+
+
+def _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters):
+    """
+    Returns fp32 state_dict reconstructed from ds checkpoint
+
+    Args:
+        - ``ds_checkpoint_dir``: path to the deepspeed checkpoint folder (where the optimizer files are)
+
+    """
+    print(f"Processing zero checkpoint '{ds_checkpoint_dir}'")
+
+    optim_files = get_optim_files(ds_checkpoint_dir)
+    zero_stage, world_size, fp32_flat_groups = parse_optim_states(optim_files, ds_checkpoint_dir)
+    print(f"Detected checkpoint of type zero stage {zero_stage}, world_size: {world_size}")
+
+    model_files = get_model_state_files(ds_checkpoint_dir)
+
+    zero_model_states = parse_model_states(model_files)
+    print(f'Parsing checkpoint created by deepspeed=={zero_model_states[0].ds_version}')
+
+    if zero_stage <= 2:
+        return _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                                          exclude_frozen_parameters)
+    elif zero_stage == 3:
+        return _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                                          exclude_frozen_parameters)
+
+
+def _zero2_merge_frozen_params(state_dict, zero_model_states):
+    if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0:
+        return
+
+    frozen_param_shapes = zero_model_states[0].frozen_param_shapes
+    frozen_param_fragments = zero_model_states[0].frozen_param_fragments
+
+    if debug:
+        num_elem = sum(s.numel() for s in frozen_param_shapes.values())
+        print(f'rank 0: {FROZEN_PARAM_SHAPES}.numel = {num_elem}')
+
+        wanted_params = len(frozen_param_shapes)
+        wanted_numel = sum(s.numel() for s in frozen_param_shapes.values())
+        avail_numel = sum([p.numel() for p in frozen_param_fragments.values()])
+        print(f'Frozen params: Have {avail_numel} numels to process.')
+        print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params')
+
+    total_params = 0
+    total_numel = 0
+    for name, shape in frozen_param_shapes.items():
+        total_params += 1
+        unpartitioned_numel = shape.numel()
+        total_numel += unpartitioned_numel
+
+        state_dict[name] = frozen_param_fragments[name]
+
+        if debug:
+            print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ")
+
+    print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _has_callable(obj, fn):
+    attr = getattr(obj, fn, None)
+    return callable(attr)
+
+
+def _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states):
+    param_shapes = zero_model_states[0].param_shapes
+
+    # Reconstruction protocol:
+    #
+    # XXX: document this
+
+    if debug:
+        for i in range(world_size):
+            for j in range(len(fp32_flat_groups[0])):
+                print(f"{FP32_FLAT_GROUPS}[{i}][{j}].shape={fp32_flat_groups[i][j].shape}")
+
+    # XXX: memory usage doubles here (zero2)
+    num_param_groups = len(fp32_flat_groups[0])
+    merged_single_partition_of_fp32_groups = []
+    for i in range(num_param_groups):
+        merged_partitions = [sd[i] for sd in fp32_flat_groups]
+        full_single_fp32_vector = torch.cat(merged_partitions, 0)
+        merged_single_partition_of_fp32_groups.append(full_single_fp32_vector)
+    avail_numel = sum(
+        [full_single_fp32_vector.numel() for full_single_fp32_vector in merged_single_partition_of_fp32_groups])
+
+    if debug:
+        wanted_params = sum([len(shapes) for shapes in param_shapes])
+        wanted_numel = sum([sum(shape.numel() for shape in shapes.values()) for shapes in param_shapes])
+        # not asserting if there is a mismatch due to possible padding
+        print(f"Have {avail_numel} numels to process.")
+        print(f"Need {wanted_numel} numels in {wanted_params} params.")
+
+    # params
+    # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support
+    # out-of-core computing solution
+    total_numel = 0
+    total_params = 0
+    for shapes, full_single_fp32_vector in zip(param_shapes, merged_single_partition_of_fp32_groups):
+        offset = 0
+        avail_numel = full_single_fp32_vector.numel()
+        for name, shape in shapes.items():
+
+            unpartitioned_numel = shape.numel() if _has_callable(shape, 'numel') else math.prod(shape)
+            total_numel += unpartitioned_numel
+            total_params += 1
+
+            if debug:
+                print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ")
+            state_dict[name] = full_single_fp32_vector.narrow(0, offset, unpartitioned_numel).view(shape)
+            offset += unpartitioned_numel
+
+        # Z2 started to align to 2*world_size to improve nccl performance. Therefore both offset and
+        # avail_numel can differ by anywhere between 0..2*world_size. Due to two unrelated complex
+        # paddings performed in the code it's almost impossible to predict the exact numbers w/o the
+        # live optimizer object, so we are checking that the numbers are within the right range
+        align_to = 2 * world_size
+
+        def zero2_align(x):
+            return align_to * math.ceil(x / align_to)
+
+        if debug:
+            print(f"original offset={offset}, avail_numel={avail_numel}")
+
+        offset = zero2_align(offset)
+        avail_numel = zero2_align(avail_numel)
+
+        if debug:
+            print(f"aligned  offset={offset}, avail_numel={avail_numel}")
+
+        # Sanity check
+        if offset != avail_numel:
+            raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong")
+
+    print(f"Reconstructed fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                               exclude_frozen_parameters):
+    state_dict = OrderedDict()
+
+    # buffers
+    buffers = zero_model_states[0].buffers
+    state_dict.update(buffers)
+    if debug:
+        print(f"added {len(buffers)} buffers")
+
+    if not exclude_frozen_parameters:
+        _zero2_merge_frozen_params(state_dict, zero_model_states)
+
+    _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states)
+
+    # recover shared parameters
+    for pair in zero_model_states[0].shared_params:
+        if pair[1] in state_dict:
+            state_dict[pair[0]] = state_dict[pair[1]]
+
+    return state_dict
+
+
+def zero3_partitioned_param_info(unpartitioned_numel, world_size):
+    remainder = unpartitioned_numel % world_size
+    padding_numel = (world_size - remainder) if remainder else 0
+    partitioned_numel = math.ceil(unpartitioned_numel / world_size)
+    return partitioned_numel, padding_numel
+
+
+def _zero3_merge_frozen_params(state_dict, world_size, zero_model_states):
+    if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0:
+        return
+
+    if debug:
+        for i in range(world_size):
+            num_elem = sum(s.numel() for s in zero_model_states[i].frozen_param_fragments.values())
+            print(f'rank {i}: {FROZEN_PARAM_SHAPES}.numel = {num_elem}')
+
+        frozen_param_shapes = zero_model_states[0].frozen_param_shapes
+        wanted_params = len(frozen_param_shapes)
+        wanted_numel = sum(s.numel() for s in frozen_param_shapes.values())
+        avail_numel = sum([p.numel() for p in zero_model_states[0].frozen_param_fragments.values()]) * world_size
+        print(f'Frozen params: Have {avail_numel} numels to process.')
+        print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params')
+
+    total_params = 0
+    total_numel = 0
+    for name, shape in zero_model_states[0].frozen_param_shapes.items():
+        total_params += 1
+        unpartitioned_numel = shape.numel()
+        total_numel += unpartitioned_numel
+
+        param_frags = tuple(model_state.frozen_param_fragments[name] for model_state in zero_model_states)
+        state_dict[name] = torch.cat(param_frags, 0).narrow(0, 0, unpartitioned_numel).view(shape)
+
+        partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size)
+
+        if debug:
+            print(
+                f"Frozen params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}"
+            )
+
+    print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states):
+    param_shapes = zero_model_states[0].param_shapes
+    avail_numel = fp32_flat_groups[0].numel() * world_size
+    # Reconstruction protocol: For zero3 we need to zip the partitions together at boundary of each
+    # param, re-consolidating each param, while dealing with padding if any
+
+    # merge list of dicts, preserving order
+    param_shapes = {k: v for d in param_shapes for k, v in d.items()}
+
+    if debug:
+        for i in range(world_size):
+            print(f"{FP32_FLAT_GROUPS}[{i}].shape={fp32_flat_groups[i].shape}")
+
+        wanted_params = len(param_shapes)
+        wanted_numel = sum(shape.numel() for shape in param_shapes.values())
+        # not asserting if there is a mismatch due to possible padding
+        avail_numel = fp32_flat_groups[0].numel() * world_size
+        print(f"Trainable params: Have {avail_numel} numels to process.")
+        print(f"Trainable params: Need {wanted_numel} numels in {wanted_params} params.")
+
+    # params
+    # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support
+    # out-of-core computing solution
+    offset = 0
+    total_numel = 0
+    total_params = 0
+    for name, shape in param_shapes.items():
+
+        unpartitioned_numel = shape.numel()
+        total_numel += unpartitioned_numel
+        total_params += 1
+
+        partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size)
+
+        if debug:
+            print(
+                f"Trainable params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}"
+            )
+
+        # XXX: memory usage doubles here
+        state_dict[name] = torch.cat(
+            tuple(fp32_flat_groups[i].narrow(0, offset, partitioned_numel) for i in range(world_size)),
+            0).narrow(0, 0, unpartitioned_numel).view(shape)
+        offset += partitioned_numel
+
+    offset *= world_size
+
+    # Sanity check
+    if offset != avail_numel:
+        raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong")
+
+    print(f"Reconstructed Trainable fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                               exclude_frozen_parameters):
+    state_dict = OrderedDict()
+
+    # buffers
+    buffers = zero_model_states[0].buffers
+    state_dict.update(buffers)
+    if debug:
+        print(f"added {len(buffers)} buffers")
+
+    if not exclude_frozen_parameters:
+        _zero3_merge_frozen_params(state_dict, world_size, zero_model_states)
+
+    _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states)
+
+    # recover shared parameters
+    for pair in zero_model_states[0].shared_params:
+        if pair[1] in state_dict:
+            state_dict[pair[0]] = state_dict[pair[1]]
+
+    return state_dict
+
+
+def get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag=None, exclude_frozen_parameters=False):
+    """
+    Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated state_dict that can be loaded with
+    ``load_state_dict()`` and used for training without DeepSpeed or shared with others, for example
+    via a model hub.
+
+    Args:
+        - ``checkpoint_dir``: path to the desired checkpoint folder
+        - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in 'latest' file. e.g., ``global_step14``
+        - ``exclude_frozen_parameters``: exclude frozen parameters
+
+    Returns:
+        - pytorch ``state_dict``
+
+    Note: this approach may not work if your application doesn't have sufficient free CPU memory and
+    you may need to use the offline approach using the ``zero_to_fp32.py`` script that is saved with
+    the checkpoint.
+
+    A typical usage might be ::
+
+        from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint
+        # do the training and checkpoint saving
+        state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir) # already on cpu
+        model = model.cpu() # move to cpu
+        model.load_state_dict(state_dict)
+        # submit to model hub or save the model to share with others
+
+    In this example the ``model`` will no longer be usable in the deepspeed context of the same
+    application. i.e. you will need to re-initialize the deepspeed engine, since
+    ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it.
+
+    If you want it all done for you, use ``load_state_dict_from_zero_checkpoint`` instead.
+
+    """
+    if tag is None:
+        latest_path = os.path.join(checkpoint_dir, 'latest')
+        if os.path.isfile(latest_path):
+            with open(latest_path, 'r') as fd:
+                tag = fd.read().strip()
+        else:
+            raise ValueError(f"Unable to find 'latest' file at {latest_path}")
+
+    ds_checkpoint_dir = os.path.join(checkpoint_dir, tag)
+
+    if not os.path.isdir(ds_checkpoint_dir):
+        raise FileNotFoundError(f"Directory '{ds_checkpoint_dir}' doesn't exist")
+
+    return _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters)
+
+
+def convert_zero_checkpoint_to_fp32_state_dict(checkpoint_dir, output_file, tag=None, exclude_frozen_parameters=False):
+    """
+    Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` file that can be
+    loaded with ``torch.load(file)`` + ``load_state_dict()`` and used for training without DeepSpeed.
+
+    Args:
+        - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``)
+        - ``output_file``: path to the pytorch fp32 state_dict output file (e.g. path/pytorch_model.bin)
+        - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14``
+        - ``exclude_frozen_parameters``: exclude frozen parameters
+    """
+
+    state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag, exclude_frozen_parameters)
+    print(f"Saving fp32 state dict to {output_file}")
+    torch.save(state_dict, output_file)
+
+
+def load_state_dict_from_zero_checkpoint(model, checkpoint_dir, tag=None):
+    """
+    1. Put the provided model to cpu
+    2. Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict``
+    3. Load it into the provided model
+
+    Args:
+        - ``model``: the model object to update
+        - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``)
+        - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14``
+
+    Returns:
+        - ``model`: modified model
+
+    Make sure you have plenty of CPU memory available before you call this function. If you don't
+    have enough use the ``zero_to_fp32.py`` utility to do the conversion. You will find it
+    conveniently placed for you in the checkpoint folder.
+
+    A typical usage might be ::
+
+        from deepspeed.utils.zero_to_fp32 import load_state_dict_from_zero_checkpoint
+        model = load_state_dict_from_zero_checkpoint(trainer.model, checkpoint_dir)
+        # submit to model hub or save the model to share with others
+
+    Note, that once this was run, the ``model`` will no longer be usable in the deepspeed context
+    of the same application. i.e. you will need to re-initialize the deepspeed engine, since
+    ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it.
+
+    """
+    logger.info(f"Extracting fp32 weights")
+    state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag)
+
+    logger.info(f"Overwriting model with fp32 weights")
+    model = model.cpu()
+    model.load_state_dict(state_dict, strict=False)
+
+    return model
+
+
+if __name__ == "__main__":
+
+    parser = argparse.ArgumentParser()
+    parser.add_argument("checkpoint_dir",
+                        type=str,
+                        help="path to the desired checkpoint folder, e.g., path/checkpoint-12")
+    parser.add_argument(
+        "output_file",
+        type=str,
+        help="path to the pytorch fp32 state_dict output file (e.g. path/checkpoint-12/pytorch_model.bin)")
+    parser.add_argument("-t",
+                        "--tag",
+                        type=str,
+                        default=None,
+                        help="checkpoint tag used as a unique identifier for checkpoint. e.g., global_step1")
+    parser.add_argument("--exclude_frozen_parameters", action='store_true', help="exclude frozen parameters")
+    parser.add_argument("-d", "--debug", action='store_true', help="enable debug")
+    args = parser.parse_args()
+
+    debug = args.debug
+
+    convert_zero_checkpoint_to_fp32_state_dict(args.checkpoint_dir,
+                                               args.output_file,
+                                               tag=args.tag,
+                                               exclude_frozen_parameters=args.exclude_frozen_parameters)