Upload 9 files

configs/410M-og-pythia-gptj-chatbbot.yml

@@ -0,0 +1,112 @@
+{
+
+  "train_data_paths": ["/workspace/harshit_data/ChatbotData/bin_data/train/bin__text_document"],
+  "valid_data_paths": ["/workspace/harshit_data/ChatbotData/bin_data/val/bin__text_document"],
+  "test_data_paths": ["/workspace/harshit_data/ChatbotData/bin_data/test/bin__text_document"],
+
+  "tokenizer_type": "HFTokenizer",
+  "vocab_file": "/workspace/harshit_data/ChatbotData/harshit_chatbot_tokenizer.json",
+
+
+  "save": "checkpoints/410M_og_pythia_gpt_j_chatbot",
+  "load": "checkpoints/410M_og_pythia_gpt_j_chatbot",
+  "tensorboard_dir": "tensorboard/410M_og_pythia_gpt_j_chatbot",
+  "log_dir": "logs/410M_og_pythia_gpt_j_chatbot",   
+
+
+  "use_wandb": False,
+  # parallelism settings
+  "pipe_parallel_size": 1,
+  "model_parallel_size": 4,
+
+  # model settings
+  "num_layers": 24,
+  "hidden_size": 1024,
+  "num_attention_heads": 16,
+  "seq_length": 2048,
+  "max_position_embeddings": 2048,
+  "pos_emb": "rotary",
+  "rotary_pct": 0.25,
+  "no_weight_tying": true,
+  "gpt_j_residual": true,
+  "output_layer_parallelism": "column",
+  
+  "attention_config": [[["flash"], 24]],
+
+  "scaled_upper_triang_masked_softmax_fusion": true,
+  "bias_gelu_fusion": true,
+
+  # init methods
+  "init_method": "small_init",
+  "output_layer_init_method": "wang_init",
+
+  "optimizer": {
+    "type": "Adam",
+    "params": {
+      "lr": 0.0003,
+      "betas": [0.9, 0.95],
+      "eps": 1.0e-8
+    }
+  },
+  "min_lr": 0.00003,
+
+  "zero_optimization": {
+    "stage": 1,
+    "allgather_partitions": true,
+    "allgather_bucket_size": 500000000,
+    "overlap_comm": true,
+    "reduce_scatter": true,
+    "reduce_bucket_size": 500000000,
+    "contiguous_gradients": true,
+    "cpu_offload": false
+  },
+
+  # batch size (trained on 32 gpus)
+  "train_micro_batch_size_per_gpu": 32,
+  "data_impl": "mmap",
+  "num_workers": 1,
+
+  # activation checkpointing
+  "checkpoint_activations": true,
+  "checkpoint_num_layers": 1,
+  "partition_activations": true,
+  "synchronize_each_layer": true,
+
+  # regularization
+  "gradient_clipping": 1.0,
+  "weight_decay": 0.1,
+  "hidden_dropout": 0,
+  "attention_dropout": 0,
+
+  # precision settings
+  "fp16": {
+    "fp16": true,
+    "enabled": true,
+    "loss_scale": 0,
+    "loss_scale_window": 1000,
+    "initial_scale_power": 12,
+    "hysteresis": 2,
+    "min_loss_scale": 1,
+  },
+
+  # misc. training settings
+   "train_iters": 100000,
+   "lr_decay_iters": 100000,
+   "distributed_backend": "nccl",
+   "lr_decay_style": "cosine",
+   "warmup": 0.00,
+   "checkpoint_factor": 5000,
+   "eval_interval": 5000,
+   "eval_iters": 10,
+   "do_test": true,
+   "extra_save_iters": [10,100,500,1000],
+
+   # logging
+   "log_interval": 100,
+   "steps_per_print": 10,
+   "keep_last_n_checkpoints": 10,
+   "wall_clock_breakdown": true,
+
+
+ 
+}

mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c6c4291dc827a50a8a20ea845c81172368ec66542d9596251ad615b8a59ee098
+size 184500188

mp_rank_01_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5d6acb47d996e5d795cb02143bfcbf7a381d7ae064c38c6599301e66190723de
+size 184500188

mp_rank_02_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:491c8dcb7d42ea18dc50b0329b4951877d38fdf0125ad60c6c1d5063f2543ae5
+size 184500188

mp_rank_03_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a17f3481845d6c342773c3dd294a5ffd3a9f6d13e63b71dcaf5cd2fd86ef2be8
+size 184500188

zero_pp_rank_0_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c884e2ec8fa83a4d45115d02f11f704a6c6dae93984f6e2ebeb38244642054c5
+size 1106487998

zero_pp_rank_0_mp_rank_01_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bcf22cacc45cd8882f734587e6cd4284861f5aa8bf41aee17bda3e6b795a855c
+size 1106487998

zero_pp_rank_0_mp_rank_02_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1e22abaddcb2b1b279db5880059c2f94da84d6ef9c0234fc013985826cc42dda
+size 1106487998

zero_pp_rank_0_mp_rank_03_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:21c84c0951d327c81fc46967a55bc9662286ee027a6637af72cf279dfb32d2c4
+size 1106487998