upload

.gitattributes

@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+tokenizer.json filter=lfs diff=lfs merge=lfs -text

added_tokens.json

@@ -0,0 +1,24 @@
+{
+  "</tool_call>": 151658,
+  "<tool_call>": 151657,
+  "<|box_end|>": 151649,
+  "<|box_start|>": 151648,
+  "<|endoftext|>": 151643,
+  "<|file_sep|>": 151664,
+  "<|fim_middle|>": 151660,
+  "<|fim_pad|>": 151662,
+  "<|fim_prefix|>": 151659,
+  "<|fim_suffix|>": 151661,
+  "<|im_end|>": 151645,
+  "<|im_start|>": 151644,
+  "<|image_pad|>": 151655,
+  "<|object_ref_end|>": 151647,
+  "<|object_ref_start|>": 151646,
+  "<|quad_end|>": 151651,
+  "<|quad_start|>": 151650,
+  "<|repo_name|>": 151663,
+  "<|video_pad|>": 151656,
+  "<|vision_end|>": 151653,
+  "<|vision_pad|>": 151654,
+  "<|vision_start|>": 151652
+}

config.json

@@ -0,0 +1,30 @@
+{
+  "_name_or_path": "/lustre/fsn1/projects/rech/gkb/uua32zb/grand_challenge/checkpoints/Qwen__Qwen2.5-1.5B-pretraining-fineweb2-0.0001LR-8192CL-1GAS-4BS-1EPOCHS-0.9BETA1-0.95BETA2/",
+  "architectures": [
+    "Qwen2ForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151643,
+  "hidden_act": "silu",
+  "hidden_size": 1536,
+  "initializer_range": 0.02,
+  "intermediate_size": 8960,
+  "max_position_embeddings": 131072,
+  "max_window_layers": 28,
+  "model_type": "qwen2",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 2,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": null,
+  "tie_word_embeddings": true,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.46.1",
+  "use_cache": false,
+  "use_mrope": false,
+  "use_sliding_window": false,
+  "vocab_size": 151936
+}

generation_config.json

@@ -0,0 +1,6 @@
+{
+  "bos_token_id": 151643,
+  "eos_token_id": 151643,
+  "max_new_tokens": 2048,
+  "transformers_version": "4.46.1"
+}

latest

@@ -0,0 +1 @@
+global_step6237

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cc5edf1bc45f08dfaeca221b40e789f61302043d25115984818a52a274f213be
+size 3554214752

special_tokens_map.json

@@ -0,0 +1,31 @@
+{
+  "additional_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "eos_token": {
+    "content": "<|im_end|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9c5ae00e602b8860cbd784ba82a8aa14e8feecec692e7076590d014d7b7fdafa
+size 11421896

tokenizer_config.json

@@ -0,0 +1,208 @@
+{
+  "add_bos_token": false,
+  "add_prefix_space": false,
+  "added_tokens_decoder": {
+    "151643": {
+      "content": "<|endoftext|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151644": {
+      "content": "<|im_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151645": {
+      "content": "<|im_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151646": {
+      "content": "<|object_ref_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151647": {
+      "content": "<|object_ref_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151648": {
+      "content": "<|box_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151649": {
+      "content": "<|box_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151650": {
+      "content": "<|quad_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151651": {
+      "content": "<|quad_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151652": {
+      "content": "<|vision_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151653": {
+      "content": "<|vision_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151654": {
+      "content": "<|vision_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151655": {
+      "content": "<|image_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151656": {
+      "content": "<|video_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151657": {
+      "content": "<tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151658": {
+      "content": "</tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151659": {
+      "content": "<|fim_prefix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151660": {
+      "content": "<|fim_middle|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151661": {
+      "content": "<|fim_suffix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151662": {
+      "content": "<|fim_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151663": {
+      "content": "<|repo_name|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151664": {
+      "content": "<|file_sep|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    }
+  },
+  "additional_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "bos_token": null,
+  "chat_template": "{%- if tools %}\n    {{- '<|im_start|>system\\n' }}\n    {%- if messages[0]['role'] == 'system' %}\n        {{- messages[0]['content'] }}\n    {%- else %}\n        {{- 'You are a helpful assistant.' }}\n    {%- endif %}\n    {{- \"\\n\\n# Tools\\n\\nYou may call one or more functions to assist with the user query.\\n\\nYou are provided with function signatures within <tools></tools> XML tags:\\n<tools>\" }}\n    {%- for tool in tools %}\n        {{- \"\\n\" }}\n        {{- tool | tojson }}\n    {%- endfor %}\n    {{- \"\\n</tools>\\n\\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\\n<tool_call>\\n{\\\"name\\\": <function-name>, \\\"arguments\\\": <args-json-object>}\\n</tool_call><|im_end|>\\n\" }}\n{%- else %}\n    {%- if messages[0]['role'] == 'system' %}\n        {{- '<|im_start|>system\\n' + messages[0]['content'] + '<|im_end|>\\n' }}\n    {%- else %}\n        {{- '<|im_start|>system\\nYou are a helpful assistant.<|im_end|>\\n' }}\n    {%- endif %}\n{%- endif %}\n{%- for message in messages %}\n    {%- if (message.role == \"user\") or (message.role == \"system\" and not loop.first) or (message.role == \"assistant\" and not message.tool_calls) %}\n        {{- '<|im_start|>' + message.role + '\\n' + message.content + '<|im_end|>' + '\\n' }}\n    {%- elif message.role == \"assistant\" %}\n        {{- '<|im_start|>' + message.role }}\n        {%- if message.content %}\n            {{- '\\n' + message.content }}\n        {%- endif %}\n        {%- for tool_call in message.tool_calls %}\n            {%- if tool_call.function is defined %}\n                {%- set tool_call = tool_call.function %}\n            {%- endif %}\n            {{- '\\n<tool_call>\\n{\"name\": \"' }}\n            {{- tool_call.name }}\n            {{- '\", \"arguments\": ' }}\n            {{- tool_call.arguments | tojson }}\n            {{- '}\\n</tool_call>' }}\n        {%- endfor %}\n        {{- '<|im_end|>\\n' }}\n    {%- elif message.role == \"tool\" %}\n        {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != \"tool\") %}\n            {{- '<|im_start|>user' }}\n        {%- endif %}\n        {{- '\\n<tool_response>\\n' }}\n        {{- message.content }}\n        {{- '\\n</tool_response>' }}\n        {%- if loop.last or (messages[loop.index0 + 1].role != \"tool\") %}\n            {{- '<|im_end|>\\n' }}\n        {%- endif %}\n    {%- endif %}\n{%- endfor %}\n{%- if add_generation_prompt %}\n    {{- '<|im_start|>assistant\\n' }}\n{%- endif %}\n",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "model_max_length": 131072,
+  "pad_token": "<|endoftext|>",
+  "padding_side": "right",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}

trainer_state.json

@@ -0,0 +1,901 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500.0,
+  "global_step": 6237,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.02405002405002405,
+      "grad_norm": 0.4139963388442993,
+      "learning_rate": 0.00019996828714700116,
+      "loss": 1.5971,
+      "step": 50
+    },
+    {
+      "epoch": 0.0481000481000481,
+      "grad_norm": 0.3423018157482147,
+      "learning_rate": 0.00019987316870210547,
+      "loss": 1.274,
+      "step": 100
+    },
+    {
+      "epoch": 0.07215007215007214,
+      "grad_norm": 0.3551710247993469,
+      "learning_rate": 0.0001997147049948582,
+      "loss": 1.2519,
+      "step": 150
+    },
+    {
+      "epoch": 0.0962000962000962,
+      "grad_norm": 0.32329073548316956,
+      "learning_rate": 0.0001994929965319844,
+      "loss": 1.2382,
+      "step": 200
+    },
+    {
+      "epoch": 0.12025012025012025,
+      "grad_norm": 0.48585018515586853,
+      "learning_rate": 0.0001992081839336419,
+      "loss": 1.2293,
+      "step": 250
+    },
+    {
+      "epoch": 0.1443001443001443,
+      "grad_norm": 0.40136224031448364,
+      "learning_rate": 0.00019886044784423197,
+      "loss": 1.2214,
+      "step": 300
+    },
+    {
+      "epoch": 0.16835016835016836,
+      "grad_norm": 0.574002206325531,
+      "learning_rate": 0.00019845000881782432,
+      "loss": 1.2184,
+      "step": 350
+    },
+    {
+      "epoch": 0.1924001924001924,
+      "grad_norm": 0.4179827570915222,
+      "learning_rate": 0.00019797712717826914,
+      "loss": 1.2064,
+      "step": 400
+    },
+    {
+      "epoch": 0.21645021645021645,
+      "grad_norm": 0.33033809065818787,
+      "learning_rate": 0.00019744210285408488,
+      "loss": 1.2055,
+      "step": 450
+    },
+    {
+      "epoch": 0.2405002405002405,
+      "grad_norm": 0.2719138562679291,
+      "learning_rate": 0.0001968452751882264,
+      "loss": 1.2077,
+      "step": 500
+    },
+    {
+      "epoch": 0.26455026455026454,
+      "grad_norm": 0.29797521233558655,
+      "learning_rate": 0.00019618702272285434,
+      "loss": 1.2096,
+      "step": 550
+    },
+    {
+      "epoch": 0.2886002886002886,
+      "grad_norm": 0.3336372673511505,
+      "learning_rate": 0.00019546776295924212,
+      "loss": 1.2072,
+      "step": 600
+    },
+    {
+      "epoch": 0.3126503126503126,
+      "grad_norm": 0.26755037903785706,
+      "learning_rate": 0.0001946879520929728,
+      "loss": 1.1974,
+      "step": 650
+    },
+    {
+      "epoch": 0.3367003367003367,
+      "grad_norm": 0.36268576979637146,
+      "learning_rate": 0.00019384808472459368,
+      "loss": 1.2045,
+      "step": 700
+    },
+    {
+      "epoch": 0.36075036075036077,
+      "grad_norm": 0.3121575713157654,
+      "learning_rate": 0.0001929486935459127,
+      "loss": 1.1889,
+      "step": 750
+    },
+    {
+      "epoch": 0.3848003848003848,
+      "grad_norm": 0.3159404993057251,
+      "learning_rate": 0.00019199034900213452,
+      "loss": 1.1921,
+      "step": 800
+    },
+    {
+      "epoch": 0.40885040885040885,
+      "grad_norm": 0.7236579060554504,
+      "learning_rate": 0.000190973658930052,
+      "loss": 1.194,
+      "step": 850
+    },
+    {
+      "epoch": 0.4329004329004329,
+      "grad_norm": 0.24907168745994568,
+      "learning_rate": 0.00018989926817252113,
+      "loss": 1.191,
+      "step": 900
+    },
+    {
+      "epoch": 0.45695045695045694,
+      "grad_norm": 0.24481187760829926,
+      "learning_rate": 0.00018876785816946505,
+      "loss": 1.1857,
+      "step": 950
+    },
+    {
+      "epoch": 0.481000481000481,
+      "grad_norm": 0.2668200731277466,
+      "learning_rate": 0.00018758014652566597,
+      "loss": 1.1957,
+      "step": 1000
+    },
+    {
+      "epoch": 0.5050505050505051,
+      "grad_norm": 0.2687171399593353,
+      "learning_rate": 0.0001863368865556191,
+      "loss": 1.1864,
+      "step": 1050
+    },
+    {
+      "epoch": 0.5291005291005291,
+      "grad_norm": 0.23915782570838928,
+      "learning_rate": 0.0001850388668057379,
+      "loss": 1.184,
+      "step": 1100
+    },
+    {
+      "epoch": 0.5531505531505532,
+      "grad_norm": 0.37159469723701477,
+      "learning_rate": 0.0001836869105542127,
+      "loss": 1.1849,
+      "step": 1150
+    },
+    {
+      "epoch": 0.5772005772005772,
+      "grad_norm": 0.2752649784088135,
+      "learning_rate": 0.0001822818752888408,
+      "loss": 1.1843,
+      "step": 1200
+    },
+    {
+      "epoch": 0.6012506012506013,
+      "grad_norm": 0.19733025133609772,
+      "learning_rate": 0.00018082465216315882,
+      "loss": 1.1766,
+      "step": 1250
+    },
+    {
+      "epoch": 0.6253006253006252,
+      "grad_norm": 0.2180165797472,
+      "learning_rate": 0.00017931616543122214,
+      "loss": 1.1865,
+      "step": 1300
+    },
+    {
+      "epoch": 0.6493506493506493,
+      "grad_norm": 0.25025510787963867,
+      "learning_rate": 0.00017775737186139038,
+      "loss": 1.1723,
+      "step": 1350
+    },
+    {
+      "epoch": 0.6734006734006734,
+      "grad_norm": 0.2865007817745209,
+      "learning_rate": 0.00017614926012949028,
+      "loss": 1.172,
+      "step": 1400
+    },
+    {
+      "epoch": 0.6974506974506974,
+      "grad_norm": 0.3406023681163788,
+      "learning_rate": 0.00017449285019174098,
+      "loss": 1.1795,
+      "step": 1450
+    },
+    {
+      "epoch": 0.7215007215007215,
+      "grad_norm": 0.19766800105571747,
+      "learning_rate": 0.00017278919263783978,
+      "loss": 1.1784,
+      "step": 1500
+    },
+    {
+      "epoch": 0.7455507455507455,
+      "grad_norm": 0.1965962052345276,
+      "learning_rate": 0.00017103936802461797,
+      "loss": 1.1754,
+      "step": 1550
+    },
+    {
+      "epoch": 0.7696007696007696,
+      "grad_norm": 0.2381555736064911,
+      "learning_rate": 0.00016924448619069023,
+      "loss": 1.1671,
+      "step": 1600
+    },
+    {
+      "epoch": 0.7936507936507936,
+      "grad_norm": 0.20156389474868774,
+      "learning_rate": 0.00016740568555253155,
+      "loss": 1.1738,
+      "step": 1650
+    },
+    {
+      "epoch": 0.8177008177008177,
+      "grad_norm": 0.18294361233711243,
+      "learning_rate": 0.00016552413238242857,
+      "loss": 1.1727,
+      "step": 1700
+    },
+    {
+      "epoch": 0.8417508417508418,
+      "grad_norm": 0.2975623309612274,
+      "learning_rate": 0.00016360102006876317,
+      "loss": 1.1677,
+      "step": 1750
+    },
+    {
+      "epoch": 0.8658008658008658,
+      "grad_norm": 0.1871371865272522,
+      "learning_rate": 0.0001616375683590974,
+      "loss": 1.1689,
+      "step": 1800
+    },
+    {
+      "epoch": 0.8898508898508899,
+      "grad_norm": 0.21457934379577637,
+      "learning_rate": 0.00015963502258654005,
+      "loss": 1.1605,
+      "step": 1850
+    },
+    {
+      "epoch": 0.9139009139009139,
+      "grad_norm": 0.20261706411838531,
+      "learning_rate": 0.0001575946528798853,
+      "loss": 1.1627,
+      "step": 1900
+    },
+    {
+      "epoch": 0.937950937950938,
+      "grad_norm": 0.17685186862945557,
+      "learning_rate": 0.0001555177533580245,
+      "loss": 1.1627,
+      "step": 1950
+    },
+    {
+      "epoch": 0.962000962000962,
+      "grad_norm": 0.212468221783638,
+      "learning_rate": 0.00015340564130914233,
+      "loss": 1.161,
+      "step": 2000
+    },
+    {
+      "epoch": 0.9860509860509861,
+      "grad_norm": 0.175174742937088,
+      "learning_rate": 0.00015125965635521724,
+      "loss": 1.1688,
+      "step": 2050
+    },
+    {
+      "epoch": 1.0101010101010102,
+      "grad_norm": 0.19970253109931946,
+      "learning_rate": 0.00014908115960235682,
+      "loss": 1.142,
+      "step": 2100
+    },
+    {
+      "epoch": 1.034151034151034,
+      "grad_norm": 0.21254608035087585,
+      "learning_rate": 0.00014687153277750676,
+      "loss": 1.1271,
+      "step": 2150
+    },
+    {
+      "epoch": 1.0582010582010581,
+      "grad_norm": 0.1651500016450882,
+      "learning_rate": 0.00014463217735208062,
+      "loss": 1.121,
+      "step": 2200
+    },
+    {
+      "epoch": 1.0822510822510822,
+      "grad_norm": 0.2405405044555664,
+      "learning_rate": 0.00014236451365306674,
+      "loss": 1.1313,
+      "step": 2250
+    },
+    {
+      "epoch": 1.1063011063011063,
+      "grad_norm": 0.17223596572875977,
+      "learning_rate": 0.00014006997996217593,
+      "loss": 1.1344,
+      "step": 2300
+    },
+    {
+      "epoch": 1.1303511303511304,
+      "grad_norm": 0.1969347894191742,
+      "learning_rate": 0.00013775003160360096,
+      "loss": 1.1176,
+      "step": 2350
+    },
+    {
+      "epoch": 1.1544011544011543,
+      "grad_norm": 0.187143936753273,
+      "learning_rate": 0.00013540614002096701,
+      "loss": 1.1322,
+      "step": 2400
+    },
+    {
+      "epoch": 1.1784511784511784,
+      "grad_norm": 0.1838238537311554,
+      "learning_rate": 0.00013303979184405826,
+      "loss": 1.1293,
+      "step": 2450
+    },
+    {
+      "epoch": 1.2025012025012025,
+      "grad_norm": 0.17928341031074524,
+      "learning_rate": 0.00013065248794591223,
+      "loss": 1.1268,
+      "step": 2500
+    },
+    {
+      "epoch": 1.2265512265512266,
+      "grad_norm": 0.2683047950267792,
+      "learning_rate": 0.00012824574249088063,
+      "loss": 1.1234,
+      "step": 2550
+    },
+    {
+      "epoch": 1.2506012506012505,
+      "grad_norm": 0.18034860491752625,
+      "learning_rate": 0.0001258210819742599,
+      "loss": 1.125,
+      "step": 2600
+    },
+    {
+      "epoch": 1.2746512746512746,
+      "grad_norm": 0.26357391476631165,
+      "learning_rate": 0.00012338004425410074,
+      "loss": 1.1217,
+      "step": 2650
+    },
+    {
+      "epoch": 1.2987012987012987,
+      "grad_norm": 0.17828579246997833,
+      "learning_rate": 0.00012092417757581085,
+      "loss": 1.1262,
+      "step": 2700
+    },
+    {
+      "epoch": 1.3227513227513228,
+      "grad_norm": 0.20247310400009155,
+      "learning_rate": 0.00011845503959016928,
+      "loss": 1.1246,
+      "step": 2750
+    },
+    {
+      "epoch": 1.3468013468013469,
+      "grad_norm": 0.17381271719932556,
+      "learning_rate": 0.0001159741963653755,
+      "loss": 1.1181,
+      "step": 2800
+    },
+    {
+      "epoch": 1.370851370851371,
+      "grad_norm": 0.19958114624023438,
+      "learning_rate": 0.00011348322139375948,
+      "loss": 1.1307,
+      "step": 2850
+    },
+    {
+      "epoch": 1.3949013949013949,
+      "grad_norm": 0.21912401914596558,
+      "learning_rate": 0.00011098369459378328,
+      "loss": 1.1264,
+      "step": 2900
+    },
+    {
+      "epoch": 1.418951418951419,
+      "grad_norm": 0.1694297194480896,
+      "learning_rate": 0.00010847720130796631,
+      "loss": 1.1256,
+      "step": 2950
+    },
+    {
+      "epoch": 1.443001443001443,
+      "grad_norm": 0.13446395099163055,
+      "learning_rate": 0.00010596533129737092,
+      "loss": 1.1258,
+      "step": 3000
+    },
+    {
+      "epoch": 1.467051467051467,
+      "grad_norm": 0.140371173620224,
+      "learning_rate": 0.00010344967773328507,
+      "loss": 1.1191,
+      "step": 3050
+    },
+    {
+      "epoch": 1.491101491101491,
+      "grad_norm": 0.18016813695430756,
+      "learning_rate": 0.00010093183618674224,
+      "loss": 1.114,
+      "step": 3100
+    },
+    {
+      "epoch": 1.5151515151515151,
+      "grad_norm": 0.17306862771511078,
+      "learning_rate": 9.84134036165192e-05,
+      "loss": 1.1149,
+      "step": 3150
+    },
+    {
+      "epoch": 1.5392015392015392,
+      "grad_norm": 0.14116255939006805,
+      "learning_rate": 9.589597735625377e-05,
+      "loss": 1.123,
+      "step": 3200
+    },
+    {
+      "epoch": 1.5632515632515633,
+      "grad_norm": 0.16819800436496735,
+      "learning_rate": 9.338115410132441e-05,
+      "loss": 1.1203,
+      "step": 3250
+    },
+    {
+      "epoch": 1.5873015873015874,
+      "grad_norm": 0.21958529949188232,
+      "learning_rate": 9.087052889613518e-05,
+      "loss": 1.1226,
+      "step": 3300
+    },
+    {
+      "epoch": 1.6113516113516113,
+      "grad_norm": 0.15786272287368774,
+      "learning_rate": 8.836569412244745e-05,
+      "loss": 1.1212,
+      "step": 3350
+    },
+    {
+      "epoch": 1.6354016354016354,
+      "grad_norm": 0.17366796731948853,
+      "learning_rate": 8.586823848940047e-05,
+      "loss": 1.1129,
+      "step": 3400
+    },
+    {
+      "epoch": 1.6594516594516593,
+      "grad_norm": 0.21448016166687012,
+      "learning_rate": 8.337974602586152e-05,
+      "loss": 1.1216,
+      "step": 3450
+    },
+    {
+      "epoch": 1.6835016835016834,
+      "grad_norm": 0.17243099212646484,
+      "learning_rate": 8.090179507574427e-05,
+      "loss": 1.1096,
+      "step": 3500
+    },
+    {
+      "epoch": 1.7075517075517075,
+      "grad_norm": 0.1429734081029892,
+      "learning_rate": 7.843595729693316e-05,
+      "loss": 1.1071,
+      "step": 3550
+    },
+    {
+      "epoch": 1.7316017316017316,
+      "grad_norm": 0.15200386941432953,
+      "learning_rate": 7.598379666444808e-05,
+      "loss": 1.1158,
+      "step": 3600
+    },
+    {
+      "epoch": 1.7556517556517557,
+      "grad_norm": 0.1442406326532364,
+      "learning_rate": 7.354686847848242e-05,
+      "loss": 1.112,
+      "step": 3650
+    },
+    {
+      "epoch": 1.7797017797017798,
+      "grad_norm": 0.17678239941596985,
+      "learning_rate": 7.11267183779428e-05,
+      "loss": 1.1118,
+      "step": 3700
+    },
+    {
+      "epoch": 1.8037518037518039,
+      "grad_norm": 0.147593155503273,
+      "learning_rate": 6.872488136011667e-05,
+      "loss": 1.1165,
+      "step": 3750
+    },
+    {
+      "epoch": 1.8278018278018278,
+      "grad_norm": 0.1334652155637741,
+      "learning_rate": 6.634288080708952e-05,
+      "loss": 1.1135,
+      "step": 3800
+    },
+    {
+      "epoch": 1.8518518518518519,
+      "grad_norm": 0.14890378713607788,
+      "learning_rate": 6.398222751952899e-05,
+      "loss": 1.1086,
+      "step": 3850
+    },
+    {
+      "epoch": 1.8759018759018757,
+      "grad_norm": 0.1334807574748993,
+      "learning_rate": 6.164441875844882e-05,
+      "loss": 1.1144,
+      "step": 3900
+    },
+    {
+      "epoch": 1.8999518999518998,
+      "grad_norm": 0.12897680699825287,
+      "learning_rate": 5.933093729556062e-05,
+      "loss": 1.1116,
+      "step": 3950
+    },
+    {
+      "epoch": 1.924001924001924,
+      "grad_norm": 0.17530564963817596,
+      "learning_rate": 5.7043250472815356e-05,
+      "loss": 1.1039,
+      "step": 4000
+    },
+    {
+      "epoch": 1.948051948051948,
+      "grad_norm": 0.15966495871543884,
+      "learning_rate": 5.478280927173145e-05,
+      "loss": 1.101,
+      "step": 4050
+    },
+    {
+      "epoch": 1.9721019721019721,
+      "grad_norm": 0.18890446424484253,
+      "learning_rate": 5.255104739309924e-05,
+      "loss": 1.1077,
+      "step": 4100
+    },
+    {
+      "epoch": 1.9961519961519962,
+      "grad_norm": 0.1547369807958603,
+      "learning_rate": 5.0349380347646494e-05,
+      "loss": 1.103,
+      "step": 4150
+    },
+    {
+      "epoch": 2.0202020202020203,
+      "grad_norm": 0.13888758420944214,
+      "learning_rate": 4.8179204558240444e-05,
+      "loss": 1.0826,
+      "step": 4200
+    },
+    {
+      "epoch": 2.0442520442520444,
+      "grad_norm": 0.11266086250543594,
+      "learning_rate": 4.6041896474197e-05,
+      "loss": 1.071,
+      "step": 4250
+    },
+    {
+      "epoch": 2.068302068302068,
+      "grad_norm": 0.14245671033859253,
+      "learning_rate": 4.393881169825779e-05,
+      "loss": 1.0759,
+      "step": 4300
+    },
+    {
+      "epoch": 2.092352092352092,
+      "grad_norm": 0.1226249411702156,
+      "learning_rate": 4.187128412678969e-05,
+      "loss": 1.0742,
+      "step": 4350
+    },
+    {
+      "epoch": 2.1164021164021163,
+      "grad_norm": 0.12307476997375488,
+      "learning_rate": 3.984062510375155e-05,
+      "loss": 1.0721,
+      "step": 4400
+    },
+    {
+      "epoch": 2.1404521404521404,
+      "grad_norm": 0.12813834846019745,
+      "learning_rate": 3.7848122588965144e-05,
+      "loss": 1.0726,
+      "step": 4450
+    },
+    {
+      "epoch": 2.1645021645021645,
+      "grad_norm": 0.13432885706424713,
+      "learning_rate": 3.5895040341217543e-05,
+      "loss": 1.0745,
+      "step": 4500
+    },
+    {
+      "epoch": 2.1885521885521886,
+      "grad_norm": 0.11649097502231598,
+      "learning_rate": 3.398261711671309e-05,
+      "loss": 1.079,
+      "step": 4550
+    },
+    {
+      "epoch": 2.2126022126022127,
+      "grad_norm": 0.11140163242816925,
+      "learning_rate": 3.211206588338358e-05,
+      "loss": 1.0748,
+      "step": 4600
+    },
+    {
+      "epoch": 2.236652236652237,
+      "grad_norm": 0.10978424549102783,
+      "learning_rate": 3.028457305155483e-05,
+      "loss": 1.0726,
+      "step": 4650
+    },
+    {
+      "epoch": 2.260702260702261,
+      "grad_norm": 0.11395589262247086,
+      "learning_rate": 2.8501297721457422e-05,
+      "loss": 1.0656,
+      "step": 4700
+    },
+    {
+      "epoch": 2.284752284752285,
+      "grad_norm": 0.10599405318498611,
+      "learning_rate": 2.6763370948059353e-05,
+      "loss": 1.0765,
+      "step": 4750
+    },
+    {
+      "epoch": 2.3088023088023086,
+      "grad_norm": 0.11157254874706268,
+      "learning_rate": 2.5071895023686442e-05,
+      "loss": 1.0726,
+      "step": 4800
+    },
+    {
+      "epoch": 2.3328523328523327,
+      "grad_norm": 0.1390163153409958,
+      "learning_rate": 2.342794277888547e-05,
+      "loss": 1.0731,
+      "step": 4850
+    },
+    {
+      "epoch": 2.356902356902357,
+      "grad_norm": 0.1519329994916916,
+      "learning_rate": 2.1832556901973965e-05,
+      "loss": 1.0704,
+      "step": 4900
+    },
+    {
+      "epoch": 2.380952380952381,
+      "grad_norm": 0.1278182566165924,
+      "learning_rate": 2.0286749277707782e-05,
+      "loss": 1.0661,
+      "step": 4950
+    },
+    {
+      "epoch": 2.405002405002405,
+      "grad_norm": 0.10508263111114502,
+      "learning_rate": 1.879150034548588e-05,
+      "loss": 1.0758,
+      "step": 5000
+    },
+    {
+      "epoch": 2.429052429052429,
+      "grad_norm": 0.09690719097852707,
+      "learning_rate": 1.7347758477500044e-05,
+      "loss": 1.0644,
+      "step": 5050
+    },
+    {
+      "epoch": 2.4531024531024532,
+      "grad_norm": 0.10174595564603806,
+      "learning_rate": 1.5956439377222798e-05,
+      "loss": 1.0726,
+      "step": 5100
+    },
+    {
+      "epoch": 2.4771524771524773,
+      "grad_norm": 0.10294167697429657,
+      "learning_rate": 1.4618425498616162e-05,
+      "loss": 1.0655,
+      "step": 5150
+    },
+    {
+      "epoch": 2.501202501202501,
+      "grad_norm": 0.11103129386901855,
+      "learning_rate": 1.3334565486428996e-05,
+      "loss": 1.0651,
+      "step": 5200
+    },
+    {
+      "epoch": 2.525252525252525,
+      "grad_norm": 0.10614852607250214,
+      "learning_rate": 1.2105673637938053e-05,
+      "loss": 1.0701,
+      "step": 5250
+    },
+    {
+      "epoch": 2.549302549302549,
+      "grad_norm": 0.09437720477581024,
+      "learning_rate": 1.0932529386474188e-05,
+      "loss": 1.0673,
+      "step": 5300
+    },
+    {
+      "epoch": 2.5733525733525733,
+      "grad_norm": 0.0965106412768364,
+      "learning_rate": 9.815876807061264e-06,
+      "loss": 1.0769,
+      "step": 5350
+    },
+    {
+      "epoch": 2.5974025974025974,
+      "grad_norm": 0.09335634112358093,
+      "learning_rate": 8.756424144481312e-06,
+      "loss": 1.0646,
+      "step": 5400
+    },
+    {
+      "epoch": 2.6214526214526215,
+      "grad_norm": 0.09890544414520264,
+      "learning_rate": 7.75484336406529e-06,
+      "loss": 1.0757,
+      "step": 5450
+    },
+    {
+      "epoch": 2.6455026455026456,
+      "grad_norm": 0.09670912474393845,
+      "learning_rate": 6.8117697254943106e-06,
+      "loss": 1.0668,
+      "step": 5500
+    },
+    {
+      "epoch": 2.6695526695526697,
+      "grad_norm": 0.09898468106985092,
+      "learning_rate": 5.927801379881714e-06,
+      "loss": 1.0745,
+      "step": 5550
+    },
+    {
+      "epoch": 2.6936026936026938,
+      "grad_norm": 0.08697386831045151,
+      "learning_rate": 5.103498990391509e-06,
+      "loss": 1.0653,
+      "step": 5600
+    },
+    {
+      "epoch": 2.717652717652718,
+      "grad_norm": 0.09457134455442429,
+      "learning_rate": 4.339385376633775e-06,
+      "loss": 1.0678,
+      "step": 5650
+    },
+    {
+      "epoch": 2.741702741702742,
+      "grad_norm": 0.09092475473880768,
+      "learning_rate": 3.6359451830626723e-06,
+      "loss": 1.0635,
+      "step": 5700
+    },
+    {
+      "epoch": 2.7657527657527656,
+      "grad_norm": 0.08736653625965118,
+      "learning_rate": 2.993624571587239e-06,
+      "loss": 1.0639,
+      "step": 5750
+    },
+    {
+      "epoch": 2.7898027898027897,
+      "grad_norm": 0.09138292819261551,
+      "learning_rate": 2.4128309385900717e-06,
+      "loss": 1.065,
+      "step": 5800
+    },
+    {
+      "epoch": 2.813852813852814,
+      "grad_norm": 0.08842656016349792,
+      "learning_rate": 1.8939326565333037e-06,
+      "loss": 1.0636,
+      "step": 5850
+    },
+    {
+      "epoch": 2.837902837902838,
+      "grad_norm": 0.08870802819728851,
+      "learning_rate": 1.437258840315714e-06,
+      "loss": 1.0706,
+      "step": 5900
+    },
+    {
+      "epoch": 2.861952861952862,
+      "grad_norm": 0.08659425377845764,
+      "learning_rate": 1.0430991385293575e-06,
+      "loss": 1.0673,
+      "step": 5950
+    },
+    {
+      "epoch": 2.886002886002886,
+      "grad_norm": 0.08142086863517761,
+      "learning_rate": 7.117035497478553e-07,
+      "loss": 1.0697,
+      "step": 6000
+    },
+    {
+      "epoch": 2.91005291005291,
+      "grad_norm": 0.080448217689991,
+      "learning_rate": 4.432822639630407e-07,
+      "loss": 1.0655,
+      "step": 6050
+    },
+    {
+      "epoch": 2.934102934102934,
+      "grad_norm": 0.08980288356542587,
+      "learning_rate": 2.380055292704575e-07,
+      "loss": 1.0701,
+      "step": 6100
+    },
+    {
+      "epoch": 2.958152958152958,
+      "grad_norm": 0.08309097588062286,
+      "learning_rate": 9.600354388833443e-08,
+      "loss": 1.0684,
+      "step": 6150
+    },
+    {
+      "epoch": 2.982202982202982,
+      "grad_norm": 0.08456841111183167,
+      "learning_rate": 1.7366373578442397e-08,
+      "loss": 1.0684,
+      "step": 6200
+    }
+  ],
+  "logging_steps": 50,
+  "max_steps": 6237,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.056700790948663e+20,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0cb09fa3cec0d925b5877a57afba4d17f256716f468a4f84dfa477dd700225e0
+size 6968

zero_to_fp32.py

@@ -0,0 +1,674 @@
+#!/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 . output_dir/
+#   or
+#   python zero_to_fp32.py . output_dir/ --safe_serialization
+
+import argparse
+import torch
+import glob
+import math
+import os
+import re
+import json
+from tqdm import tqdm
+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 tqdm(param_shapes.items(), desc='Gathering Sharded Weights'):
+        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_dir,
+                                               max_shard_size="5GB",
+                                               safe_serialization=False,
+                                               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_dir``: directory to the pytorch fp32 state_dict output files
+        - ``max_shard_size``: the maximum size for a checkpoint before being sharded, default value is 5GB
+        - ``safe_serialization``:  whether to save the model using `safetensors` or the traditional PyTorch way (that uses `pickle`).
+        - ``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
+    """
+    # Dependency pre-check
+    if safe_serialization:
+        try:
+            from safetensors.torch import save_file
+        except ImportError:
+            print('If you want to use `safe_serialization`, please `pip install safetensors`')
+            raise
+    if max_shard_size is not None:
+        try:
+            from huggingface_hub import split_torch_state_dict_into_shards
+        except ImportError:
+            print('If you want to use `max_shard_size`, please `pip install huggingface_hub`')
+            raise
+
+    # Convert zero checkpoint to state_dict
+    state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag, exclude_frozen_parameters)
+
+    # Shard the model if it is too big.
+    weights_name = "model.safetensors" if safe_serialization else "pytorch_model.bin"
+    if max_shard_size is not None:
+        filename_pattern = weights_name.replace(".bin", "{suffix}.bin").replace(".safetensors", "{suffix}.safetensors")
+        state_dict_split = split_torch_state_dict_into_shards(state_dict,
+                                                              filename_pattern=filename_pattern,
+                                                              max_shard_size=max_shard_size)
+    else:
+        from collections import namedtuple
+        StateDictSplit = namedtuple("StateDictSplit", ["is_sharded", "filename_to_tensors"])
+        state_dict_split = StateDictSplit(is_sharded=False,
+                                          filename_to_tensors={weights_name: list(state_dict.keys())})
+
+    # Save the model
+    filename_to_tensors = state_dict_split.filename_to_tensors.items()
+    for shard_file, tensors in tqdm(filename_to_tensors, desc="Saving checkpoint shards"):
+        shard = {tensor: state_dict[tensor].contiguous() for tensor in tensors}
+        output_path = os.path.join(output_dir, shard_file)
+        if safe_serialization:
+            save_file(shard, output_path, metadata={"format": "pt"})
+        else:
+            torch.save(shard, output_path)
+
+    # Save index if sharded
+    if state_dict_split.is_sharded:
+        index = {
+            "metadata": state_dict_split.metadata,
+            "weight_map": state_dict_split.tensor_to_filename,
+        }
+        save_index_file = "model.safetensors.index.json" if safe_serialization else "pytorch_model.bin.index.json"
+        save_index_file = os.path.join(output_dir, save_index_file)
+        with open(save_index_file, "w", encoding="utf-8") as f:
+            content = json.dumps(index, indent=2, sort_keys=True) + "\n"
+            f.write(content)
+
+
+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_dir",
+                        type=str,
+                        help="directory to the pytorch fp32 state_dict output files"
+                        "(e.g. path/checkpoint-12-output/)")
+    parser.add_argument(
+        "--max_shard_size",
+        type=str,
+        default="5GB",
+        help="The maximum size for a checkpoint before being sharded. Checkpoints shard will then be each of size"
+        "lower than this size. If expressed as a string, needs to be digits followed by a unit (like `5MB`"
+        "We default it to 5GB in order for models to be able to run easily on free-tier google colab instances"
+        "without CPU OOM issues.")
+    parser.add_argument(
+        "--safe_serialization",
+        default=False,
+        action='store_true',
+        help="Whether to save the model using `safetensors` or the traditional PyTorch way (that uses `pickle`).")
+    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_dir,
+                                               max_shard_size=args.max_shard_size,
+                                               safe_serialization=args.safe_serialization,
+                                               tag=args.tag,
+                                               exclude_frozen_parameters=args.exclude_frozen_parameters)