Upload folder using huggingface_hub

111_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": [
+    "o_proj",
+    "v_proj",
+    "gate_proj",
+    "down_proj",
+    "k_proj",
+    "q_proj",
+    "up_proj"
+  ],
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_dora": false,
+  "use_rslora": false
+}

111_128_e3_3e-5/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a511ab5e01ce1b601dc329bfe1ce2d61c8f9a1d838ffbbab2c306b397938a2c1
+size 791751704

111_128_e3_3e-5/latest

@@ -0,0 +1 @@
+global_step537

111_128_e3_3e-5/rng_state_0.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:114ff88107235766cd246023a5b06ef976c8d689addb328a75d642f0d8dd4705
+size 16389

111_128_e3_3e-5/rng_state_1.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8f41b251f15eaa032bb3c909554fce47ae23b252d6d568b4b2e6a84d33b7f0d4
+size 16389

111_128_e3_3e-5/rng_state_2.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:927088adf9ae95ce55c553259b18d31bc3cfa16399afb11a57bb999581908d9d
+size 16389

111_128_e3_3e-5/rng_state_3.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c3c0276893be2edbe9e10065cbdba3bb1ffc5f33f236859c06496fce3a0ede69
+size 16389

111_128_e3_3e-5/rng_state_4.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:edcd0cac16ca9effb055e79aacc4f5d4bdb962d67d0844cc8c704eae93fcb9e8
+size 16389

111_128_e3_3e-5/rng_state_5.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f9bd6c677d133fbe382c2be384bd9ffacec40ab535610edd2fe5377f72e393ad
+size 16389

111_128_e3_3e-5/rng_state_6.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:707e5711b5437c739d9e4c2f97bf85d7ff7ccc2489225e9a7a8129bc70ee6621
+size 16389

111_128_e3_3e-5/rng_state_7.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:af4a518eb061371fe44fcc7c6056023e605a56db6faa2674ad3fb03104372668
+size 16389

111_128_e3_3e-5/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ac92e68338b30a0259d238704f80ace9c61ecadb37745f28cd992dee7b518398
+size 1401

111_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
+  }
+}

111_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
+}

111_128_e3_3e-5/trainer_state.json

@@ -0,0 +1,783 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 537,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.028011204481792718,
+      "grad_norm": 0.8302801251411438,
+      "learning_rate": 4.444444444444444e-06,
+      "loss": 1.2795,
+      "step": 5
+    },
+    {
+      "epoch": 0.056022408963585436,
+      "grad_norm": 0.7679424285888672,
+      "learning_rate": 9.999999999999999e-06,
+      "loss": 1.3132,
+      "step": 10
+    },
+    {
+      "epoch": 0.08403361344537816,
+      "grad_norm": 0.6632059812545776,
+      "learning_rate": 1.5555555555555555e-05,
+      "loss": 1.2617,
+      "step": 15
+    },
+    {
+      "epoch": 0.11204481792717087,
+      "grad_norm": 0.5141933560371399,
+      "learning_rate": 2.111111111111111e-05,
+      "loss": 1.2446,
+      "step": 20
+    },
+    {
+      "epoch": 0.1400560224089636,
+      "grad_norm": 0.5640107989311218,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 1.2056,
+      "step": 25
+    },
+    {
+      "epoch": 0.16806722689075632,
+      "grad_norm": 0.6174699664115906,
+      "learning_rate": 2.999886165172246e-05,
+      "loss": 1.1521,
+      "step": 30
+    },
+    {
+      "epoch": 0.19607843137254902,
+      "grad_norm": 0.5704578161239624,
+      "learning_rate": 2.99860572177674e-05,
+      "loss": 1.1697,
+      "step": 35
+    },
+    {
+      "epoch": 0.22408963585434175,
+      "grad_norm": 0.5647729635238647,
+      "learning_rate": 2.9959037600786822e-05,
+      "loss": 1.0866,
+      "step": 40
+    },
+    {
+      "epoch": 0.25210084033613445,
+      "grad_norm": 0.563290536403656,
+      "learning_rate": 2.99178284305241e-05,
+      "loss": 1.0377,
+      "step": 45
+    },
+    {
+      "epoch": 0.2801120448179272,
+      "grad_norm": 0.6452869176864624,
+      "learning_rate": 2.9862468796373404e-05,
+      "loss": 1.0318,
+      "step": 50
+    },
+    {
+      "epoch": 0.3081232492997199,
+      "grad_norm": 0.6674671769142151,
+      "learning_rate": 2.9793011210301036e-05,
+      "loss": 0.9634,
+      "step": 55
+    },
+    {
+      "epoch": 0.33613445378151263,
+      "grad_norm": 0.5875696539878845,
+      "learning_rate": 2.9709521557034668e-05,
+      "loss": 0.9359,
+      "step": 60
+    },
+    {
+      "epoch": 0.3641456582633053,
+      "grad_norm": 0.6587890386581421,
+      "learning_rate": 2.9612079031567654e-05,
+      "loss": 0.8687,
+      "step": 65
+    },
+    {
+      "epoch": 0.39215686274509803,
+      "grad_norm": 0.8074948787689209,
+      "learning_rate": 2.9500776064037813e-05,
+      "loss": 0.8456,
+      "step": 70
+    },
+    {
+      "epoch": 0.42016806722689076,
+      "grad_norm": 0.7204945087432861,
+      "learning_rate": 2.9375718232051835e-05,
+      "loss": 0.8113,
+      "step": 75
+    },
+    {
+      "epoch": 0.4481792717086835,
+      "grad_norm": 0.8563284277915955,
+      "learning_rate": 2.923702416053852e-05,
+      "loss": 0.8216,
+      "step": 80
+    },
+    {
+      "epoch": 0.47619047619047616,
+      "grad_norm": 0.7782193422317505,
+      "learning_rate": 2.908482540922585e-05,
+      "loss": 0.7487,
+      "step": 85
+    },
+    {
+      "epoch": 0.5042016806722689,
+      "grad_norm": 0.8910718560218811,
+      "learning_rate": 2.891926634784862e-05,
+      "loss": 0.72,
+      "step": 90
+    },
+    {
+      "epoch": 0.5322128851540616,
+      "grad_norm": 0.9688562154769897,
+      "learning_rate": 2.8740504019205006e-05,
+      "loss": 0.7164,
+      "step": 95
+    },
+    {
+      "epoch": 0.5602240896358543,
+      "grad_norm": 1.1969927549362183,
+      "learning_rate": 2.8548707990191933e-05,
+      "loss": 0.6634,
+      "step": 100
+    },
+    {
+      "epoch": 0.5882352941176471,
+      "grad_norm": 1.052741289138794,
+      "learning_rate": 2.8344060190960646e-05,
+      "loss": 0.6355,
+      "step": 105
+    },
+    {
+      "epoch": 0.6162464985994398,
+      "grad_norm": 1.04100501537323,
+      "learning_rate": 2.812675474234489e-05,
+      "loss": 0.6794,
+      "step": 110
+    },
+    {
+      "epoch": 0.6442577030812325,
+      "grad_norm": 1.1578627824783325,
+      "learning_rate": 2.7896997771725588e-05,
+      "loss": 0.6195,
+      "step": 115
+    },
+    {
+      "epoch": 0.6722689075630253,
+      "grad_norm": 1.084123134613037,
+      "learning_rate": 2.76550072175065e-05,
+      "loss": 0.5657,
+      "step": 120
+    },
+    {
+      "epoch": 0.7002801120448179,
+      "grad_norm": 1.0797803401947021,
+      "learning_rate": 2.7401012622386454e-05,
+      "loss": 0.5493,
+      "step": 125
+    },
+    {
+      "epoch": 0.7282913165266106,
+      "grad_norm": 1.21173894405365,
+      "learning_rate": 2.7135254915624213e-05,
+      "loss": 0.532,
+      "step": 130
+    },
+    {
+      "epoch": 0.7563025210084033,
+      "grad_norm": 1.1563632488250732,
+      "learning_rate": 2.6857986184502452e-05,
+      "loss": 0.5201,
+      "step": 135
+    },
+    {
+      "epoch": 0.7843137254901961,
+      "grad_norm": 1.1629258394241333,
+      "learning_rate": 2.6569469435207712e-05,
+      "loss": 0.427,
+      "step": 140
+    },
+    {
+      "epoch": 0.8123249299719888,
+      "grad_norm": 1.1777229309082031,
+      "learning_rate": 2.6269978343353103e-05,
+      "loss": 0.4561,
+      "step": 145
+    },
+    {
+      "epoch": 0.8403361344537815,
+      "grad_norm": 1.2430548667907715,
+      "learning_rate": 2.5959796994380397e-05,
+      "loss": 0.4719,
+      "step": 150
+    },
+    {
+      "epoch": 0.8683473389355743,
+      "grad_norm": 1.237862467765808,
+      "learning_rate": 2.5639219614087803e-05,
+      "loss": 0.4232,
+      "step": 155
+    },
+    {
+      "epoch": 0.896358543417367,
+      "grad_norm": 1.1241263151168823,
+      "learning_rate": 2.530855028953894e-05,
+      "loss": 0.4175,
+      "step": 160
+    },
+    {
+      "epoch": 0.9243697478991597,
+      "grad_norm": 1.3301092386245728,
+      "learning_rate": 2.496810268061787e-05,
+      "loss": 0.3822,
+      "step": 165
+    },
+    {
+      "epoch": 0.9523809523809523,
+      "grad_norm": 1.100689172744751,
+      "learning_rate": 2.4618199722503676e-05,
+      "loss": 0.431,
+      "step": 170
+    },
+    {
+      "epoch": 0.9803921568627451,
+      "grad_norm": 1.0635671615600586,
+      "learning_rate": 2.4259173319346894e-05,
+      "loss": 0.3604,
+      "step": 175
+    },
+    {
+      "epoch": 1.0056022408963585,
+      "grad_norm": 1.2241755723953247,
+      "learning_rate": 2.3891364029438323e-05,
+      "loss": 0.3668,
+      "step": 180
+    },
+    {
+      "epoch": 1.0336134453781514,
+      "grad_norm": 1.127846121788025,
+      "learning_rate": 2.351512074216885e-05,
+      "loss": 0.2703,
+      "step": 185
+    },
+    {
+      "epoch": 1.061624649859944,
+      "grad_norm": 1.340333104133606,
+      "learning_rate": 2.313080034708674e-05,
+      "loss": 0.2839,
+      "step": 190
+    },
+    {
+      "epoch": 1.0896358543417366,
+      "grad_norm": 1.1510874032974243,
+      "learning_rate": 2.273876739536627e-05,
+      "loss": 0.2958,
+      "step": 195
+    },
+    {
+      "epoch": 1.1176470588235294,
+      "grad_norm": 1.2898104190826416,
+      "learning_rate": 2.2339393754008854e-05,
+      "loss": 0.2699,
+      "step": 200
+    },
+    {
+      "epoch": 1.145658263305322,
+      "grad_norm": 1.1437090635299683,
+      "learning_rate": 2.19330582531047e-05,
+      "loss": 0.2734,
+      "step": 205
+    },
+    {
+      "epoch": 1.173669467787115,
+      "grad_norm": 1.2426291704177856,
+      "learning_rate": 2.1520146326489476e-05,
+      "loss": 0.2598,
+      "step": 210
+    },
+    {
+      "epoch": 1.2016806722689075,
+      "grad_norm": 1.1473838090896606,
+      "learning_rate": 2.1101049646137008e-05,
+      "loss": 0.2494,
+      "step": 215
+    },
+    {
+      "epoch": 1.2296918767507004,
+      "grad_norm": 1.2547862529754639,
+      "learning_rate": 2.0676165750634656e-05,
+      "loss": 0.2078,
+      "step": 220
+    },
+    {
+      "epoch": 1.257703081232493,
+      "grad_norm": 1.1667771339416504,
+      "learning_rate": 2.0245897668093917e-05,
+      "loss": 0.2104,
+      "step": 225
+    },
+    {
+      "epoch": 1.2857142857142856,
+      "grad_norm": 1.591986060142517,
+      "learning_rate": 1.9810653533853826e-05,
+      "loss": 0.2343,
+      "step": 230
+    },
+    {
+      "epoch": 1.3137254901960784,
+      "grad_norm": 1.2143700122833252,
+      "learning_rate": 1.937084620333987e-05,
+      "loss": 0.2018,
+      "step": 235
+    },
+    {
+      "epoch": 1.3417366946778713,
+      "grad_norm": 1.4105778932571411,
+      "learning_rate": 1.8926892860445607e-05,
+      "loss": 0.2166,
+      "step": 240
+    },
+    {
+      "epoch": 1.3697478991596639,
+      "grad_norm": 1.0881307125091553,
+      "learning_rate": 1.847921462180847e-05,
+      "loss": 0.2081,
+      "step": 245
+    },
+    {
+      "epoch": 1.3977591036414565,
+      "grad_norm": 1.5105844736099243,
+      "learning_rate": 1.8028236137355154e-05,
+      "loss": 0.2817,
+      "step": 250
+    },
+    {
+      "epoch": 1.4257703081232493,
+      "grad_norm": 1.2569842338562012,
+      "learning_rate": 1.7574385187495396e-05,
+      "loss": 0.2134,
+      "step": 255
+    },
+    {
+      "epoch": 1.453781512605042,
+      "grad_norm": 1.258494257926941,
+      "learning_rate": 1.7118092277346372e-05,
+      "loss": 0.1755,
+      "step": 260
+    },
+    {
+      "epoch": 1.4817927170868348,
+      "grad_norm": 1.1781706809997559,
+      "learning_rate": 1.6659790228372512e-05,
+      "loss": 0.1829,
+      "step": 265
+    },
+    {
+      "epoch": 1.5098039215686274,
+      "grad_norm": 1.0851811170578003,
+      "learning_rate": 1.6199913767828126e-05,
+      "loss": 0.1949,
+      "step": 270
+    },
+    {
+      "epoch": 1.53781512605042,
+      "grad_norm": 1.1227320432662964,
+      "learning_rate": 1.5738899116392254e-05,
+      "loss": 0.1581,
+      "step": 275
+    },
+    {
+      "epoch": 1.5658263305322129,
+      "grad_norm": 1.225533127784729,
+      "learning_rate": 1.5277183574386947e-05,
+      "loss": 0.1578,
+      "step": 280
+    },
+    {
+      "epoch": 1.5938375350140057,
+      "grad_norm": 1.2944084405899048,
+      "learning_rate": 1.4815205106971424e-05,
+      "loss": 0.1691,
+      "step": 285
+    },
+    {
+      "epoch": 1.6218487394957983,
+      "grad_norm": 1.0687532424926758,
+      "learning_rate": 1.435340192870557e-05,
+      "loss": 0.1633,
+      "step": 290
+    },
+    {
+      "epoch": 1.649859943977591,
+      "grad_norm": 1.1152864694595337,
+      "learning_rate": 1.3892212087876892e-05,
+      "loss": 0.1456,
+      "step": 295
+    },
+    {
+      "epoch": 1.6778711484593838,
+      "grad_norm": 1.0214799642562866,
+      "learning_rate": 1.3432073050985201e-05,
+      "loss": 0.1559,
+      "step": 300
+    },
+    {
+      "epoch": 1.7058823529411766,
+      "grad_norm": 1.0343576669692993,
+      "learning_rate": 1.297342128777911e-05,
+      "loss": 0.1425,
+      "step": 305
+    },
+    {
+      "epoch": 1.7338935574229692,
+      "grad_norm": 1.3290977478027344,
+      "learning_rate": 1.251669185723805e-05,
+      "loss": 0.1601,
+      "step": 310
+    },
+    {
+      "epoch": 1.7619047619047619,
+      "grad_norm": 1.2549246549606323,
+      "learning_rate": 1.2062317994892499e-05,
+      "loss": 0.1262,
+      "step": 315
+    },
+    {
+      "epoch": 1.7899159663865545,
+      "grad_norm": 1.2381296157836914,
+      "learning_rate": 1.1610730701873788e-05,
+      "loss": 0.1205,
+      "step": 320
+    },
+    {
+      "epoch": 1.8179271708683473,
+      "grad_norm": 1.0439088344573975,
+      "learning_rate": 1.1162358336083466e-05,
+      "loss": 0.1385,
+      "step": 325
+    },
+    {
+      "epoch": 1.8459383753501402,
+      "grad_norm": 1.089748501777649,
+      "learning_rate": 1.071762620586987e-05,
+      "loss": 0.1248,
+      "step": 330
+    },
+    {
+      "epoch": 1.8739495798319328,
+      "grad_norm": 1.1243932247161865,
+      "learning_rate": 1.0276956166597384e-05,
+      "loss": 0.1327,
+      "step": 335
+    },
+    {
+      "epoch": 1.9019607843137254,
+      "grad_norm": 1.0884898900985718,
+      "learning_rate": 9.840766220491078e-06,
+      "loss": 0.127,
+      "step": 340
+    },
+    {
+      "epoch": 1.9299719887955182,
+      "grad_norm": 0.9122500419616699,
+      "learning_rate": 9.40947012013629e-06,
+      "loss": 0.1193,
+      "step": 345
+    },
+    {
+      "epoch": 1.957983193277311,
+      "grad_norm": 0.96335768699646,
+      "learning_rate": 8.98347697600922e-06,
+      "loss": 0.1081,
+      "step": 350
+    },
+    {
+      "epoch": 1.9859943977591037,
+      "grad_norm": 0.9969227910041809,
+      "learning_rate": 8.563190868410867e-06,
+      "loss": 0.1026,
+      "step": 355
+    },
+    {
+      "epoch": 2.011204481792717,
+      "grad_norm": 0.7521359920501709,
+      "learning_rate": 8.149010464172392e-06,
+      "loss": 0.1018,
+      "step": 360
+    },
+    {
+      "epoch": 2.0392156862745097,
+      "grad_norm": 0.849561333656311,
+      "learning_rate": 7.741328638495477e-06,
+      "loss": 0.0828,
+      "step": 365
+    },
+    {
+      "epoch": 2.0672268907563027,
+      "grad_norm": 0.9802676439285278,
+      "learning_rate": 7.340532102286399e-06,
+      "loss": 0.0771,
+      "step": 370
+    },
+    {
+      "epoch": 2.0952380952380953,
+      "grad_norm": 1.0678410530090332,
+      "learning_rate": 6.947001035337299e-06,
+      "loss": 0.0886,
+      "step": 375
+    },
+    {
+      "epoch": 2.123249299719888,
+      "grad_norm": 0.7994552850723267,
+      "learning_rate": 6.561108725702653e-06,
+      "loss": 0.0747,
+      "step": 380
+    },
+    {
+      "epoch": 2.1512605042016806,
+      "grad_norm": 0.5987146496772766,
+      "learning_rate": 6.1832212156129045e-06,
+      "loss": 0.0783,
+      "step": 385
+    },
+    {
+      "epoch": 2.179271708683473,
+      "grad_norm": 0.9091212153434753,
+      "learning_rate": 5.813696954261253e-06,
+      "loss": 0.093,
+      "step": 390
+    },
+    {
+      "epoch": 2.2072829131652663,
+      "grad_norm": 0.7514382004737854,
+      "learning_rate": 5.452886457792882e-06,
+      "loss": 0.0842,
+      "step": 395
+    },
+    {
+      "epoch": 2.235294117647059,
+      "grad_norm": 0.7744276523590088,
+      "learning_rate": 5.101131976819165e-06,
+      "loss": 0.0849,
+      "step": 400
+    },
+    {
+      "epoch": 2.2633053221288515,
+      "grad_norm": 0.7660306096076965,
+      "learning_rate": 4.7587671717722184e-06,
+      "loss": 0.0836,
+      "step": 405
+    },
+    {
+      "epoch": 2.291316526610644,
+      "grad_norm": 0.6748756170272827,
+      "learning_rate": 4.426116796407794e-06,
+      "loss": 0.0724,
+      "step": 410
+    },
+    {
+      "epoch": 2.3193277310924367,
+      "grad_norm": 0.6959658265113831,
+      "learning_rate": 4.10349638975664e-06,
+      "loss": 0.0764,
+      "step": 415
+    },
+    {
+      "epoch": 2.34733893557423,
+      "grad_norm": 0.7949005961418152,
+      "learning_rate": 3.791211976816634e-06,
+      "loss": 0.0763,
+      "step": 420
+    },
+    {
+      "epoch": 2.3753501400560224,
+      "grad_norm": 0.7894933223724365,
+      "learning_rate": 3.489559778269545e-06,
+      "loss": 0.0708,
+      "step": 425
+    },
+    {
+      "epoch": 2.403361344537815,
+      "grad_norm": 0.696286678314209,
+      "learning_rate": 3.198825929497752e-06,
+      "loss": 0.08,
+      "step": 430
+    },
+    {
+      "epoch": 2.431372549019608,
+      "grad_norm": 0.7180067300796509,
+      "learning_rate": 2.919286209167511e-06,
+      "loss": 0.0707,
+      "step": 435
+    },
+    {
+      "epoch": 2.4593837535014007,
+      "grad_norm": 0.6607317328453064,
+      "learning_rate": 2.6512057776361935e-06,
+      "loss": 0.0746,
+      "step": 440
+    },
+    {
+      "epoch": 2.4873949579831933,
+      "grad_norm": 0.6334755420684814,
+      "learning_rate": 2.394838925431611e-06,
+      "loss": 0.0616,
+      "step": 445
+    },
+    {
+      "epoch": 2.515406162464986,
+      "grad_norm": 0.6440517902374268,
+      "learning_rate": 2.1504288320420613e-06,
+      "loss": 0.0649,
+      "step": 450
+    },
+    {
+      "epoch": 2.5434173669467786,
+      "grad_norm": 0.7231543660163879,
+      "learning_rate": 1.9182073352458375e-06,
+      "loss": 0.0644,
+      "step": 455
+    },
+    {
+      "epoch": 2.571428571428571,
+      "grad_norm": 0.6651074290275574,
+      "learning_rate": 1.6983947111990717e-06,
+      "loss": 0.0794,
+      "step": 460
+    },
+    {
+      "epoch": 2.5994397759103642,
+      "grad_norm": 0.7134129405021667,
+      "learning_rate": 1.4911994654904404e-06,
+      "loss": 0.0643,
+      "step": 465
+    },
+    {
+      "epoch": 2.627450980392157,
+      "grad_norm": 0.8511032462120056,
+      "learning_rate": 1.2968181353609854e-06,
+      "loss": 0.0718,
+      "step": 470
+    },
+    {
+      "epoch": 2.6554621848739495,
+      "grad_norm": 0.699251651763916,
+      "learning_rate": 1.115435103276657e-06,
+      "loss": 0.0733,
+      "step": 475
+    },
+    {
+      "epoch": 2.6834733893557425,
+      "grad_norm": 0.6353307366371155,
+      "learning_rate": 9.472224220303427e-07,
+      "loss": 0.0567,
+      "step": 480
+    },
+    {
+      "epoch": 2.711484593837535,
+      "grad_norm": 0.5931615829467773,
+      "learning_rate": 7.923396515393966e-07,
+      "loss": 0.0694,
+      "step": 485
+    },
+    {
+      "epoch": 2.7394957983193278,
+      "grad_norm": 0.7215616106987,
+      "learning_rate": 6.509337074933891e-07,
+      "loss": 0.059,
+      "step": 490
+    },
+    {
+      "epoch": 2.7675070028011204,
+      "grad_norm": 0.7181894183158875,
+      "learning_rate": 5.231387219956645e-07,
+      "loss": 0.076,
+      "step": 495
+    },
+    {
+      "epoch": 2.795518207282913,
+      "grad_norm": 0.6250295639038086,
+      "learning_rate": 4.090759163309282e-07,
+      "loss": 0.0682,
+      "step": 500
+    },
+    {
+      "epoch": 2.8235294117647056,
+      "grad_norm": 0.5698807239532471,
+      "learning_rate": 3.088534859795117e-07,
+      "loss": 0.0592,
+      "step": 505
+    },
+    {
+      "epoch": 2.8515406162464987,
+      "grad_norm": 0.7776156067848206,
+      "learning_rate": 2.2256649798740204e-07,
+      "loss": 0.0715,
+      "step": 510
+    },
+    {
+      "epoch": 2.8795518207282913,
+      "grad_norm": 0.6085256338119507,
+      "learning_rate": 1.5029680078939567e-07,
+      "loss": 0.0706,
+      "step": 515
+    },
+    {
+      "epoch": 2.907563025210084,
+      "grad_norm": 0.6178684830665588,
+      "learning_rate": 9.211294657089587e-08,
+      "loss": 0.0604,
+      "step": 520
+    },
+    {
+      "epoch": 2.935574229691877,
+      "grad_norm": 0.4936996400356293,
+      "learning_rate": 4.807012624201468e-08,
+      "loss": 0.0547,
+      "step": 525
+    },
+    {
+      "epoch": 2.9635854341736696,
+      "grad_norm": 0.6194062829017639,
+      "learning_rate": 1.8210117085651902e-08,
+      "loss": 0.0722,
+      "step": 530
+    },
+    {
+      "epoch": 2.991596638655462,
+      "grad_norm": 0.6318842172622681,
+      "learning_rate": 2.561243129209356e-09,
+      "loss": 0.0626,
+      "step": 535
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 537,
+  "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": 6.756316886335488e+17,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}

111_128_e3_3e-5/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c819fd69cb57164bc47ce8e4dc62551146e0ac1d44d04f10742e829904b52848
+size 8209

111_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)