Running 1 job
{
    "type": "diffusion_trainer",
    "training_folder": "/workspace/ai-toolkit/output",
    "sqlite_db_path": "/workspace/ai-toolkit/aitk_db.db",
    "device": "cuda",
    "trigger_word": null,
    "performance_log_every": 10,
    "network": {
        "type": "lora",
        "linear": 16,
        "linear_alpha": 16,
        "conv": 16,
        "conv_alpha": 16,
        "lokr_full_rank": true,
        "lokr_factor": -1,
        "network_kwargs": {
            "ignore_if_contains": []
        }
    },
    "save": {
        "dtype": "bf16",
        "save_every": 330,
        "max_step_saves_to_keep": 8,
        "save_format": "diffusers",
        "push_to_hub": false
    },
    "datasets": [
        {
            "folder_path": "/workspace/ai-toolkit/datasets/anha",
            "mask_path": null,
            "mask_min_value": 0.1,
            "default_caption": "",
            "caption_ext": "txt",
            "caption_dropout_rate": 0.05,
            "cache_latents_to_disk": false,
            "is_reg": false,
            "network_weight": 1,
            "resolution": [
                512,
                768,
                1024
            ],
            "controls": [],
            "shrink_video_to_frames": true,
            "num_frames": 1,
            "do_i2v": true,
            "flip_x": false,
            "flip_y": false
        }
    ],
    "train": {
        "batch_size": 1,
        "bypass_guidance_embedding": false,
        "steps": 3300,
        "gradient_accumulation": 1,
        "train_unet": true,
        "train_text_encoder": false,
        "gradient_checkpointing": true,
        "noise_scheduler": "flowmatch",
        "optimizer": "adamw8bit",
        "timestep_type": "sigmoid",
        "content_or_style": "balanced",
        "optimizer_params": {
            "weight_decay": 0.0001
        },
        "unload_text_encoder": false,
        "cache_text_embeddings": false,
        "lr": 0.0001,
        "ema_config": {
            "use_ema": false,
            "ema_decay": 0.99
        },
        "skip_first_sample": false,
        "force_first_sample": false,
        "disable_sampling": false,
        "dtype": "bf16",
        "diff_output_preservation": false,
        "diff_output_preservation_multiplier": 1,
        "diff_output_preservation_class": "person",
        "switch_boundary_every": 1,
        "loss_type": "mse"
    },
    "model": {
        "name_or_path": "ai-toolkit/Wan2.2-T2V-A14B-Diffusers-bf16",
        "quantize": true,
        "qtype": "qfloat8",
        "quantize_te": true,
        "qtype_te": "qfloat8",
        "arch": "wan22_14b:t2v",
        "low_vram": true,
        "model_kwargs": {
            "train_high_noise": false,
            "train_low_noise": true
        },
        "layer_offloading": false,
        "layer_offloading_text_encoder_percent": 1,
        "layer_offloading_transformer_percent": 1
    },
    "sample": {
        "sampler": "flowmatch",
        "sample_every": 250,
        "width": 1024,
        "height": 1024,
        "samples": [],
        "neg": "",
        "seed": 42,
        "walk_seed": true,
        "guidance_scale": 4,
        "sample_steps": 25,
        "num_frames": 1,
        "fps": 16
    }
}
Using SQLite database at /workspace/ai-toolkit/aitk_db.db
Job ID: "82931103-4e27-4d03-b60f-1ea7f035166f"

#############################################
# Running job: anha
#############################################


Running  1 process
Loading Wan model
Loading transformer 1

Loading checkpoint shards:   0%|          | 0/3 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/3 [00:00<?, ?it/s]
Loading checkpoint shards:  67%|######6   | 2/3 [00:00<00:00, 14.46it/s]
Loading checkpoint shards:  67%|######6   | 2/3 [00:00<00:00, 14.46it/s]
Loading checkpoint shards: 100%|##########| 3/3 [00:00<00:00, 15.18it/s]
Loading checkpoint shards: 100%|##########| 3/3 [00:00<00:00, 15.18it/s]

Quantizing Transformer 1
 - quantizing 40 transformer blocks

  0%|          | 0/40 [00:00<?, ?it/s]
  0%|          | 0/40 [00:00<?, ?it/s]
  2%|2         | 1/40 [00:00<00:24,  1.59it/s]
  2%|2         | 1/40 [00:00<00:24,  1.59it/s]
  5%|5         | 2/40 [00:01<00:20,  1.83it/s]
  5%|5         | 2/40 [00:01<00:20,  1.83it/s]
  8%|7         | 3/40 [00:01<00:19,  1.91it/s]
  8%|7         | 3/40 [00:01<00:19,  1.91it/s]
 10%|#         | 4/40 [00:02<00:18,  1.96it/s]
 10%|#         | 4/40 [00:02<00:18,  1.96it/s]
 12%|#2        | 5/40 [00:02<00:17,  2.01it/s]
 12%|#2        | 5/40 [00:02<00:17,  2.01it/s]
 15%|#5        | 6/40 [00:03<00:16,  2.02it/s]
 15%|#5        | 6/40 [00:03<00:16,  2.02it/s]
 18%|#7        | 7/40 [00:03<00:16,  2.02it/s]
 18%|#7        | 7/40 [00:03<00:16,  2.02it/s]
 20%|##        | 8/40 [00:04<00:15,  2.07it/s]
 20%|##        | 8/40 [00:04<00:15,  2.07it/s]
 22%|##2       | 9/40 [00:04<00:14,  2.07it/s]
 22%|##2       | 9/40 [00:04<00:14,  2.07it/s]
 25%|##5       | 10/40 [00:05<00:14,  2.04it/s]
 25%|##5       | 10/40 [00:05<00:14,  2.04it/s]
 28%|##7       | 11/40 [00:05<00:14,  2.04it/s]
 28%|##7       | 11/40 [00:05<00:14,  2.04it/s]
 30%|###       | 12/40 [00:06<00:13,  2.03it/s]
 30%|###       | 12/40 [00:06<00:13,  2.03it/s]
 32%|###2      | 13/40 [00:06<00:13,  2.03it/s]
 32%|###2      | 13/40 [00:06<00:13,  2.03it/s]
 35%|###5      | 14/40 [00:07<00:13,  1.98it/s]
 35%|###5      | 14/40 [00:07<00:13,  1.98it/s]
 38%|###7      | 15/40 [00:07<00:12,  1.99it/s]
 38%|###7      | 15/40 [00:07<00:12,  1.99it/s]
 40%|####      | 16/40 [00:08<00:11,  2.01it/s]
 40%|####      | 16/40 [00:08<00:11,  2.01it/s]
 42%|####2     | 17/40 [00:08<00:11,  2.01it/s]
 42%|####2     | 17/40 [00:08<00:11,  2.01it/s]
 45%|####5     | 18/40 [00:08<00:10,  2.02it/s]
 45%|####5     | 18/40 [00:08<00:10,  2.02it/s]
 48%|####7     | 19/40 [00:09<00:10,  2.04it/s]
 48%|####7     | 19/40 [00:09<00:10,  2.04it/s]
 50%|#####     | 20/40 [00:09<00:09,  2.03it/s]
 50%|#####     | 20/40 [00:09<00:09,  2.03it/s]
 52%|#####2    | 21/40 [00:10<00:09,  2.04it/s]
 52%|#####2    | 21/40 [00:10<00:09,  2.04it/s]
 55%|#####5    | 22/40 [00:10<00:08,  2.00it/s]
 55%|#####5    | 22/40 [00:10<00:08,  2.00it/s]
 57%|#####7    | 23/40 [00:11<00:08,  2.00it/s]
 57%|#####7    | 23/40 [00:11<00:08,  2.00it/s]
 60%|######    | 24/40 [00:11<00:07,  2.01it/s]
 60%|######    | 24/40 [00:11<00:07,  2.01it/s]
 62%|######2   | 25/40 [00:12<00:07,  2.02it/s]
 62%|######2   | 25/40 [00:12<00:07,  2.02it/s]
 65%|######5   | 26/40 [00:12<00:06,  2.00it/s]
 65%|######5   | 26/40 [00:12<00:06,  2.00it/s]
 68%|######7   | 27/40 [00:13<00:06,  2.01it/s]
 68%|######7   | 27/40 [00:13<00:06,  2.01it/s]
 70%|#######   | 28/40 [00:14<00:06,  1.93it/s]
 70%|#######   | 28/40 [00:14<00:06,  1.93it/s]
 72%|#######2  | 29/40 [00:14<00:05,  1.97it/s]
 72%|#######2  | 29/40 [00:14<00:05,  1.97it/s]
 75%|#######5  | 30/40 [00:15<00:05,  1.97it/s]
 75%|#######5  | 30/40 [00:15<00:05,  1.97it/s]
 78%|#######7  | 31/40 [00:15<00:04,  2.03it/s]
 78%|#######7  | 31/40 [00:15<00:04,  2.03it/s]
 80%|########  | 32/40 [00:15<00:03,  2.03it/s]
 80%|########  | 32/40 [00:15<00:03,  2.03it/s]
 82%|########2 | 33/40 [00:16<00:03,  2.04it/s]
 82%|########2 | 33/40 [00:16<00:03,  2.04it/s]
 85%|########5 | 34/40 [00:16<00:02,  2.04it/s]
 85%|########5 | 34/40 [00:16<00:02,  2.04it/s]
 88%|########7 | 35/40 [00:17<00:02,  2.05it/s]
 88%|########7 | 35/40 [00:17<00:02,  2.05it/s]
 90%|######### | 36/40 [00:17<00:01,  2.04it/s]
 90%|######### | 36/40 [00:17<00:01,  2.04it/s]
 92%|#########2| 37/40 [00:18<00:01,  2.03it/s]
 92%|#########2| 37/40 [00:18<00:01,  2.03it/s]
 95%|#########5| 38/40 [00:18<00:00,  2.03it/s]
 95%|#########5| 38/40 [00:18<00:00,  2.03it/s]
 98%|#########7| 39/40 [00:19<00:00,  2.05it/s]
 98%|#########7| 39/40 [00:19<00:00,  2.05it/s]
100%|##########| 40/40 [00:19<00:00,  2.05it/s]
100%|##########| 40/40 [00:19<00:00,  2.05it/s]
100%|##########| 40/40 [00:19<00:00,  2.01it/s]
100%|##########| 40/40 [00:19<00:00,  2.01it/s]

 - quantizing extras
Moving transformer 1 to CPU
Loading transformer 2

Loading checkpoint shards:   0%|          | 0/3 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/3 [00:00<?, ?it/s]
Loading checkpoint shards: 100%|##########| 3/3 [00:00<00:00, 28.53it/s]
Loading checkpoint shards: 100%|##########| 3/3 [00:00<00:00, 28.53it/s]
Loading checkpoint shards: 100%|##########| 3/3 [00:00<00:00, 28.49it/s]
Loading checkpoint shards: 100%|##########| 3/3 [00:00<00:00, 28.49it/s]

Quantizing Transformer 2
 - quantizing 40 transformer blocks

  0%|          | 0/40 [00:00<?, ?it/s]
  0%|          | 0/40 [00:00<?, ?it/s]
  2%|2         | 1/40 [00:00<00:19,  2.01it/s]
  2%|2         | 1/40 [00:00<00:19,  2.01it/s]
  5%|5         | 2/40 [00:00<00:18,  2.04it/s]
  5%|5         | 2/40 [00:00<00:18,  2.04it/s]
  8%|7         | 3/40 [00:01<00:18,  2.05it/s]
  8%|7         | 3/40 [00:01<00:18,  2.05it/s]
 10%|#         | 4/40 [00:01<00:17,  2.02it/s]
 10%|#         | 4/40 [00:01<00:17,  2.02it/s]
 12%|#2        | 5/40 [00:02<00:17,  2.03it/s]
 12%|#2        | 5/40 [00:02<00:17,  2.03it/s]
 15%|#5        | 6/40 [00:02<00:16,  2.04it/s]
 15%|#5        | 6/40 [00:02<00:16,  2.04it/s]
 18%|#7        | 7/40 [00:03<00:16,  2.04it/s]
 18%|#7        | 7/40 [00:03<00:16,  2.04it/s]
 20%|##        | 8/40 [00:03<00:12,  2.52it/s]
 20%|##        | 8/40 [00:03<00:12,  2.52it/s]
 22%|##2       | 9/40 [00:03<00:10,  2.98it/s]
 22%|##2       | 9/40 [00:03<00:10,  2.98it/s]
 25%|##5       | 10/40 [00:04<00:08,  3.36it/s]
 25%|##5       | 10/40 [00:04<00:08,  3.36it/s]
 28%|##7       | 11/40 [00:04<00:08,  3.59it/s]
 28%|##7       | 11/40 [00:04<00:08,  3.59it/s]
 30%|###       | 12/40 [00:04<00:07,  3.90it/s]
 30%|###       | 12/40 [00:04<00:07,  3.90it/s]
 32%|###2      | 13/40 [00:04<00:06,  3.97it/s]
 32%|###2      | 13/40 [00:04<00:06,  3.97it/s]
 35%|###5      | 14/40 [00:05<00:07,  3.28it/s]
 35%|###5      | 14/40 [00:05<00:07,  3.28it/s]
 38%|###7      | 15/40 [00:05<00:08,  3.08it/s]
 38%|###7      | 15/40 [00:05<00:08,  3.08it/s]
 40%|####      | 16/40 [00:06<00:09,  2.64it/s]
 40%|####      | 16/40 [00:06<00:09,  2.64it/s]
 42%|####2     | 17/40 [00:06<00:09,  2.53it/s]
 42%|####2     | 17/40 [00:06<00:09,  2.53it/s]
 45%|####5     | 18/40 [00:06<00:08,  2.55it/s]
 45%|####5     | 18/40 [00:06<00:08,  2.55it/s]
 48%|####7     | 19/40 [00:07<00:08,  2.37it/s]
 48%|####7     | 19/40 [00:07<00:08,  2.37it/s]
 50%|#####     | 20/40 [00:07<00:08,  2.30it/s]
 50%|#####     | 20/40 [00:07<00:08,  2.30it/s]
 52%|#####2    | 21/40 [00:08<00:08,  2.13it/s]
 52%|#####2    | 21/40 [00:08<00:08,  2.13it/s]
 55%|#####5    | 22/40 [00:09<00:09,  1.92it/s]
 55%|#####5    | 22/40 [00:09<00:09,  1.92it/s]
 57%|#####7    | 23/40 [00:10<00:13,  1.26it/s]
 57%|#####7    | 23/40 [00:10<00:13,  1.26it/s]
 60%|######    | 24/40 [00:11<00:11,  1.36it/s]
 60%|######    | 24/40 [00:11<00:11,  1.36it/s]
 62%|######2   | 25/40 [00:11<00:09,  1.53it/s]
 62%|######2   | 25/40 [00:11<00:09,  1.53it/s]
 65%|######5   | 26/40 [00:11<00:07,  1.78it/s]
 65%|######5   | 26/40 [00:11<00:07,  1.78it/s]
 68%|######7   | 27/40 [00:12<00:06,  2.12it/s]
 68%|######7   | 27/40 [00:12<00:06,  2.12it/s]
 70%|#######   | 28/40 [00:12<00:06,  1.93it/s]
 70%|#######   | 28/40 [00:12<00:06,  1.93it/s]
 72%|#######2  | 29/40 [00:13<00:05,  1.96it/s]
 72%|#######2  | 29/40 [00:13<00:05,  1.96it/s]
 75%|#######5  | 30/40 [00:13<00:05,  1.97it/s]
 75%|#######5  | 30/40 [00:13<00:05,  1.97it/s]
 78%|#######7  | 31/40 [00:14<00:04,  1.95it/s]
 78%|#######7  | 31/40 [00:14<00:04,  1.95it/s]
 80%|########  | 32/40 [00:14<00:04,  1.98it/s]
 80%|########  | 32/40 [00:14<00:04,  1.98it/s]
 82%|########2 | 33/40 [00:15<00:03,  1.99it/s]
 82%|########2 | 33/40 [00:15<00:03,  1.99it/s]
 85%|########5 | 34/40 [00:15<00:03,  2.00it/s]
 85%|########5 | 34/40 [00:15<00:03,  2.00it/s]
 88%|########7 | 35/40 [00:16<00:02,  1.97it/s]
 88%|########7 | 35/40 [00:16<00:02,  1.97it/s]
 90%|######### | 36/40 [00:16<00:02,  1.87it/s]
 90%|######### | 36/40 [00:16<00:02,  1.87it/s]
 92%|#########2| 37/40 [00:17<00:01,  1.81it/s]
 92%|#########2| 37/40 [00:17<00:01,  1.81it/s]
 95%|#########5| 38/40 [00:17<00:01,  1.87it/s]
 95%|#########5| 38/40 [00:17<00:01,  1.87it/s]
 98%|#########7| 39/40 [00:18<00:00,  1.95it/s]
 98%|#########7| 39/40 [00:18<00:00,  1.95it/s]
100%|##########| 40/40 [00:18<00:00,  1.96it/s]
100%|##########| 40/40 [00:18<00:00,  1.96it/s]
100%|##########| 40/40 [00:18<00:00,  2.12it/s]
100%|##########| 40/40 [00:18<00:00,  2.12it/s]

 - quantizing extras
Moving transformer 2 to CPU
Creating DualWanTransformer3DModel
Loading UMT5EncoderModel
You set `add_prefix_space`. The tokenizer needs to be converted from the slow tokenizers
You set `add_prefix_space`. The tokenizer needs to be converted from the slow tokenizers
Using ai-toolkit/umt5_xxl_encoder for UMT5 encoder.

Loading checkpoint shards:   0%|          | 0/3 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/3 [00:00<?, ?it/s]
Loading checkpoint shards: 100%|##########| 3/3 [00:00<00:00, 47.28it/s]
Loading checkpoint shards: 100%|##########| 3/3 [00:00<00:00, 47.28it/s]

Quantizing UMT5EncoderModel
Moving transformer back to GPU
Loading VAE
Making pipe
Preparing Model
create LoRA network. base dim (rank): 16, alpha: 16
neuron dropout: p=None, rank dropout: p=None, module dropout: p=None
apply LoRA to Conv2d with kernel size (3,3). dim (rank): 16, alpha: 16
create LoRA for Text Encoder: 0 modules.
create LoRA for U-Net: 400 modules.
enable LoRA for U-Net
#### IMPORTANT RESUMING FROM /workspace/ai-toolkit/output/anha/anha_000001980_low_noise.safetensors ####
Loading from /workspace/ai-toolkit/output/anha/anha_000001980_low_noise.safetensors
Missing keys: []
Found step 1980 in metadata, starting from there
Loading optimizer state from /workspace/ai-toolkit/output/anha/optimizer.pt
Updating optimizer LR from params
Dataset: /workspace/ai-toolkit/datasets/anha
  -  Preprocessing image dimensions

  0%|          | 0/22 [00:00<?, ?it/s]
  0%|          | 0/22 [00:00<?, ?it/s]
100%|##########| 22/22 [00:00<00:00, 66289.29it/s]
100%|##########| 22/22 [00:00<00:00, 66289.29it/s]

  -  Found 22 images
Bucket sizes for /workspace/ai-toolkit/datasets/anha:
608x416: 3 files
416x576: 2 files
672x352: 1 files
416x608: 4 files
640x384: 1 files
512x512: 5 files
448x576: 2 files
480x544: 1 files
448x544: 1 files
544x448: 2 files
10 buckets made
Dataset: /workspace/ai-toolkit/datasets/anha
  -  Preprocessing image dimensions

  0%|          | 0/22 [00:00<?, ?it/s]
  0%|          | 0/22 [00:00<?, ?it/s]
100%|##########| 22/22 [00:00<00:00, 109589.89it/s]
100%|##########| 22/22 [00:00<00:00, 109589.89it/s]

  -  Found 22 images
Bucket sizes for /workspace/ai-toolkit/datasets/anha:
912x624: 3 files
624x864: 2 files
1008x528: 1 files
624x912: 4 files
960x576: 1 files
768x768: 5 files
672x864: 2 files
720x816: 1 files
672x816: 1 files
816x672: 2 files
10 buckets made
Dataset: /workspace/ai-toolkit/datasets/anha
  -  Preprocessing image dimensions

  0%|          | 0/22 [00:00<?, ?it/s]
  0%|          | 0/22 [00:00<?, ?it/s]
100%|##########| 22/22 [00:00<00:00, 107671.75it/s]
100%|##########| 22/22 [00:00<00:00, 107671.75it/s]

  -  Found 22 images
Bucket sizes for /workspace/ai-toolkit/datasets/anha:
1216x832: 3 files
832x1152: 2 files
1232x640: 1 files
832x1216: 4 files
1248x752: 1 files
1024x1024: 5 files
896x1152: 2 files
960x1088: 1 files
896x1088: 1 files
1088x896: 2 files
10 buckets made

anha:  60%|######    | 1980/3300 [00:00<?, ?it/s]
anha:  60%|######    | 1980/3300 [00:00<?, ?it/s]
anha:  60%|######    | 1980/3300 [00:02<?, ?it/s, lr: 1.0e-04 loss: 1.237e-01]
anha:  60%|######    | 1980/3300 [00:02<?, ?it/s, lr: 1.0e-04 loss: 1.237e-01]
anha:  60%|######    | 1980/3300 [00:02<?, ?it/s, lr: 1.0e-04 loss: 1.237e-01]
anha:  60%|######    | 1980/3300 [00:02<?, ?it/s, lr: 1.0e-04 loss: 1.237e-01]
anha:  60%|######    | 1980/3300 [00:03<?, ?it/s, lr: 1.0e-04 loss: 9.693e-02]
anha:  60%|######    | 1980/3300 [00:03<?, ?it/s, lr: 1.0e-04 loss: 9.693e-02]
anha:  60%|######    | 1981/3300 [00:03<37:36,  1.71s/it, lr: 1.0e-04 loss: 9.693e-02]
anha:  60%|######    | 1981/3300 [00:03<37:36,  1.71s/it, lr: 1.0e-04 loss: 9.693e-02]
anha:  60%|######    | 1981/3300 [00:04<37:36,  1.71s/it, lr: 1.0e-04 loss: 4.161e-02]
anha:  60%|######    | 1981/3300 [00:04<37:36,  1.71s/it, lr: 1.0e-04 loss: 4.161e-02]
anha:  60%|######    | 1982/3300 [00:04<28:27,  1.30s/it, lr: 1.0e-04 loss: 4.161e-02]
anha:  60%|######    | 1982/3300 [00:04<28:27,  1.30s/it, lr: 1.0e-04 loss: 4.161e-02]
anha:  60%|######    | 1982/3300 [00:06<28:27,  1.30s/it, lr: 1.0e-04 loss: 6.046e-02]
anha:  60%|######    | 1982/3300 [00:06<28:27,  1.30s/it, lr: 1.0e-04 loss: 6.046e-02]
anha:  60%|######    | 1983/3300 [00:06<35:11,  1.60s/it, lr: 1.0e-04 loss: 6.046e-02]
anha:  60%|######    | 1983/3300 [00:06<35:11,  1.60s/it, lr: 1.0e-04 loss: 6.046e-02]
anha:  60%|######    | 1983/3300 [00:10<35:11,  1.60s/it, lr: 1.0e-04 loss: 3.601e-02]
anha:  60%|######    | 1983/3300 [00:10<35:11,  1.60s/it, lr: 1.0e-04 loss: 3.601e-02]
anha:  60%|######    | 1984/3300 [00:10<50:20,  2.30s/it, lr: 1.0e-04 loss: 3.601e-02]
anha:  60%|######    | 1984/3300 [00:10<50:20,  2.30s/it, lr: 1.0e-04 loss: 3.601e-02]
anha:  60%|######    | 1984/3300 [00:12<50:20,  2.30s/it, lr: 1.0e-04 loss: 3.665e-02]
anha:  60%|######    | 1984/3300 [00:12<50:20,  2.30s/it, lr: 1.0e-04 loss: 3.665e-02]
anha:  60%|######    | 1985/3300 [00:12<47:33,  2.17s/it, lr: 1.0e-04 loss: 3.665e-02]
anha:  60%|######    | 1985/3300 [00:12<47:33,  2.17s/it, lr: 1.0e-04 loss: 3.665e-02]
anha:  60%|######    | 1985/3300 [00:13<47:33,  2.17s/it, lr: 1.0e-04 loss: 5.945e-02]
anha:  60%|######    | 1985/3300 [00:13<47:33,  2.17s/it, lr: 1.0e-04 loss: 5.945e-02]
anha:  60%|######    | 1986/3300 [00:13<38:46,  1.77s/it, lr: 1.0e-04 loss: 5.945e-02]
anha:  60%|######    | 1986/3300 [00:13<38:46,  1.77s/it, lr: 1.0e-04 loss: 5.945e-02]
anha:  60%|######    | 1986/3300 [00:16<38:46,  1.77s/it, lr: 1.0e-04 loss: 5.782e-02]
anha:  60%|######    | 1986/3300 [00:16<38:46,  1.77s/it, lr: 1.0e-04 loss: 5.782e-02]
anha:  60%|######    | 1987/3300 [00:16<50:07,  2.29s/it, lr: 1.0e-04 loss: 5.782e-02]
anha:  60%|######    | 1987/3300 [00:16<50:07,  2.29s/it, lr: 1.0e-04 loss: 5.782e-02]
anha:  60%|######    | 1987/3300 [00:19<50:07,  2.29s/it, lr: 1.0e-04 loss: 6.451e-02]
anha:  60%|######    | 1987/3300 [00:19<50:07,  2.29s/it, lr: 1.0e-04 loss: 6.451e-02]
anha:  60%|######    | 1988/3300 [00:19<58:16,  2.66s/it, lr: 1.0e-04 loss: 6.451e-02]
anha:  60%|######    | 1988/3300 [00:19<58:16,  2.66s/it, lr: 1.0e-04 loss: 6.451e-02]
anha:  60%|######    | 1988/3300 [00:21<58:16,  2.66s/it, lr: 1.0e-04 loss: 5.364e-02]
anha:  60%|######    | 1988/3300 [00:21<58:16,  2.66s/it, lr: 1.0e-04 loss: 5.364e-02]
anha:  60%|######    | 1989/3300 [00:21<53:30,  2.45s/it, lr: 1.0e-04 loss: 5.364e-02]
anha:  60%|######    | 1989/3300 [00:21<53:30,  2.45s/it, lr: 1.0e-04 loss: 5.364e-02]
anha:  60%|######    | 1989/3300 [00:23<53:30,  2.45s/it, lr: 1.0e-04 loss: 1.036e-01]
anha:  60%|######    | 1989/3300 [00:23<53:30,  2.45s/it, lr: 1.0e-04 loss: 1.036e-01]
anha:  60%|######    | 1990/3300 [00:24<53:27,  2.45s/it, lr: 1.0e-04 loss: 1.084e-01]
anha:  60%|######    | 1990/3300 [00:24<53:27,  2.45s/it, lr: 1.0e-04 loss: 1.084e-01]
anha:  60%|######    | 1991/3300 [00:24<33:17,  1.53s/it, lr: 1.0e-04 loss: 1.084e-01]
anha:  60%|######    | 1991/3300 [00:24<33:17,  1.53s/it, lr: 1.0e-04 loss: 1.084e-01]
anha:  60%|######    | 1991/3300 [00:25<33:17,  1.53s/it, lr: 1.0e-04 loss: 9.665e-02]
anha:  60%|######    | 1991/3300 [00:25<33:17,  1.53s/it, lr: 1.0e-04 loss: 9.665e-02]
anha:  60%|######    | 1992/3300 [00:25<30:20,  1.39s/it, lr: 1.0e-04 loss: 9.665e-02]
anha:  60%|######    | 1992/3300 [00:25<30:20,  1.39s/it, lr: 1.0e-04 loss: 9.665e-02]
anha:  60%|######    | 1992/3300 [00:26<30:20,  1.39s/it, lr: 1.0e-04 loss: 7.573e-02]
anha:  60%|######    | 1992/3300 [00:26<30:20,  1.39s/it, lr: 1.0e-04 loss: 7.573e-02]
anha:  60%|######    | 1993/3300 [00:26<28:05,  1.29s/it, lr: 1.0e-04 loss: 7.573e-02]
anha:  60%|######    | 1993/3300 [00:26<28:05,  1.29s/it, lr: 1.0e-04 loss: 7.573e-02]
anha:  60%|######    | 1993/3300 [00:30<28:05,  1.29s/it, lr: 1.0e-04 loss: 5.139e-02]
anha:  60%|######    | 1993/3300 [00:30<28:05,  1.29s/it, lr: 1.0e-04 loss: 5.139e-02]
anha:  60%|######    | 1994/3300 [00:30<41:05,  1.89s/it, lr: 1.0e-04 loss: 5.139e-02]
anha:  60%|######    | 1994/3300 [00:30<41:05,  1.89s/it, lr: 1.0e-04 loss: 5.139e-02]
anha:  60%|######    | 1994/3300 [00:33<41:05,  1.89s/it, lr: 1.0e-04 loss: 8.511e-02]
anha:  60%|######    | 1994/3300 [00:33<41:05,  1.89s/it, lr: 1.0e-04 loss: 8.511e-02]
anha:  60%|######    | 1995/3300 [00:33<49:34,  2.28s/it, lr: 1.0e-04 loss: 8.511e-02]
anha:  60%|######    | 1995/3300 [00:33<49:34,  2.28s/it, lr: 1.0e-04 loss: 8.511e-02]
anha:  60%|######    | 1995/3300 [00:37<49:34,  2.28s/it, lr: 1.0e-04 loss: 1.246e-01]
anha:  60%|######    | 1995/3300 [00:37<49:34,  2.28s/it, lr: 1.0e-04 loss: 1.246e-01]
anha:  60%|######    | 1996/3300 [00:37<56:59,  2.62s/it, lr: 1.0e-04 loss: 1.246e-01]
anha:  60%|######    | 1996/3300 [00:37<56:59,  2.62s/it, lr: 1.0e-04 loss: 1.246e-01]
anha:  60%|######    | 1996/3300 [00:39<56:59,  2.62s/it, lr: 1.0e-04 loss: 3.722e-02]
anha:  60%|######    | 1996/3300 [00:39<56:59,  2.62s/it, lr: 1.0e-04 loss: 3.722e-02]
anha:  61%|######    | 1997/3300 [00:39<52:50,  2.43s/it, lr: 1.0e-04 loss: 3.722e-02]
anha:  61%|######    | 1997/3300 [00:39<52:50,  2.43s/it, lr: 1.0e-04 loss: 3.722e-02]
anha:  61%|######    | 1997/3300 [00:40<52:50,  2.43s/it, lr: 1.0e-04 loss: 4.014e-02]
anha:  61%|######    | 1997/3300 [00:40<52:50,  2.43s/it, lr: 1.0e-04 loss: 4.014e-02]
anha:  61%|######    | 1998/3300 [00:40<43:42,  2.01s/it, lr: 1.0e-04 loss: 4.014e-02]
anha:  61%|######    | 1998/3300 [00:40<43:42,  2.01s/it, lr: 1.0e-04 loss: 4.014e-02]
anha:  61%|######    | 1998/3300 [00:41<43:42,  2.01s/it, lr: 1.0e-04 loss: 7.887e-02]
anha:  61%|######    | 1998/3300 [00:41<43:42,  2.01s/it, lr: 1.0e-04 loss: 7.887e-02]
anha:  61%|######    | 1999/3300 [00:41<37:49,  1.74s/it, lr: 1.0e-04 loss: 7.887e-02]
anha:  61%|######    | 1999/3300 [00:41<37:49,  1.74s/it, lr: 1.0e-04 loss: 7.887e-02]
anha:  61%|######    | 1999/3300 [00:44<37:49,  1.74s/it, lr: 1.0e-04 loss: 1.638e-01]
anha:  61%|######    | 1999/3300 [00:44<37:49,  1.74s/it, lr: 1.0e-04 loss: 1.638e-01]


Generating Images: 0it [00:00, ?it/s]
Generating Images: 0it [00:00, ?it/s]


                                     
                                     
anha:  61%|######    | 2000/3300 [00:46<37:47,  1.74s/it, lr: 1.0e-04 loss: 1.296e-01]
anha:  61%|######    | 2000/3300 [00:46<37:47,  1.74s/it, lr: 1.0e-04 loss: 1.296e-01]
anha:  61%|######    | 2001/3300 [00:46<30:12,  1.40s/it, lr: 1.0e-04 loss: 1.296e-01]
anha:  61%|######    | 2001/3300 [00:46<30:12,  1.40s/it, lr: 1.0e-04 loss: 1.296e-01]
anha:  61%|######    | 2001/3300 [00:47<30:12,  1.40s/it, lr: 1.0e-04 loss: 5.126e-02]
anha:  61%|######    | 2001/3300 [00:47<30:12,  1.40s/it, lr: 1.0e-04 loss: 5.126e-02]
anha:  61%|######    | 2002/3300 [00:47<28:10,  1.30s/it, lr: 1.0e-04 loss: 5.126e-02]
anha:  61%|######    | 2002/3300 [00:47<28:10,  1.30s/it, lr: 1.0e-04 loss: 5.126e-02]
anha:  61%|######    | 2002/3300 [00:50<28:10,  1.30s/it, lr: 1.0e-04 loss: 4.245e-02]
anha:  61%|######    | 2002/3300 [00:50<28:10,  1.30s/it, lr: 1.0e-04 loss: 4.245e-02]
anha:  61%|######    | 2003/3300 [00:50<40:24,  1.87s/it, lr: 1.0e-04 loss: 4.245e-02]
anha:  61%|######    | 2003/3300 [00:50<40:24,  1.87s/it, lr: 1.0e-04 loss: 4.245e-02]
anha:  61%|######    | 2003/3300 [00:54<40:24,  1.87s/it, lr: 1.0e-04 loss: 4.192e-02]
anha:  61%|######    | 2003/3300 [00:54<40:24,  1.87s/it, lr: 1.0e-04 loss: 4.192e-02]
anha:  61%|######    | 2004/3300 [00:54<48:23,  2.24s/it, lr: 1.0e-04 loss: 4.192e-02]
anha:  61%|######    | 2004/3300 [00:54<48:23,  2.24s/it, lr: 1.0e-04 loss: 4.192e-02]
anha:  61%|######    | 2004/3300 [00:55<48:23,  2.24s/it, lr: 1.0e-04 loss: 9.221e-02]
anha:  61%|######    | 2004/3300 [00:55<48:23,  2.24s/it, lr: 1.0e-04 loss: 9.221e-02]
anha:  61%|######    | 2005/3300 [00:55<40:56,  1.90s/it, lr: 1.0e-04 loss: 9.221e-02]
anha:  61%|######    | 2005/3300 [00:55<40:56,  1.90s/it, lr: 1.0e-04 loss: 9.221e-02]
anha:  61%|######    | 2005/3300 [00:56<40:56,  1.90s/it, lr: 1.0e-04 loss: 5.338e-02]
anha:  61%|######    | 2005/3300 [00:56<40:56,  1.90s/it, lr: 1.0e-04 loss: 5.338e-02]
anha:  61%|######    | 2006/3300 [00:56<35:25,  1.64s/it, lr: 1.0e-04 loss: 5.338e-02]
anha:  61%|######    | 2006/3300 [00:56<35:25,  1.64s/it, lr: 1.0e-04 loss: 5.338e-02]
anha:  61%|######    | 2006/3300 [00:57<35:25,  1.64s/it, lr: 1.0e-04 loss: 4.842e-02]
anha:  61%|######    | 2006/3300 [00:57<35:25,  1.64s/it, lr: 1.0e-04 loss: 4.842e-02]
anha:  61%|######    | 2007/3300 [00:57<31:25,  1.46s/it, lr: 1.0e-04 loss: 4.842e-02]
anha:  61%|######    | 2007/3300 [00:57<31:25,  1.46s/it, lr: 1.0e-04 loss: 4.842e-02]
anha:  61%|######    | 2007/3300 [01:00<31:25,  1.46s/it, lr: 1.0e-04 loss: 1.022e-01]
anha:  61%|######    | 2007/3300 [01:00<31:25,  1.46s/it, lr: 1.0e-04 loss: 1.022e-01]
anha:  61%|######    | 2008/3300 [01:00<43:27,  2.02s/it, lr: 1.0e-04 loss: 1.022e-01]
anha:  61%|######    | 2008/3300 [01:00<43:27,  2.02s/it, lr: 1.0e-04 loss: 1.022e-01]
anha:  61%|######    | 2008/3300 [01:01<43:27,  2.02s/it, lr: 1.0e-04 loss: 9.690e-02]
anha:  61%|######    | 2008/3300 [01:01<43:27,  2.02s/it, lr: 1.0e-04 loss: 9.690e-02]
anha:  61%|######    | 2009/3300 [01:01<37:00,  1.72s/it, lr: 1.0e-04 loss: 9.690e-02]
anha:  61%|######    | 2009/3300 [01:01<37:00,  1.72s/it, lr: 1.0e-04 loss: 9.690e-02]
anha:  61%|######    | 2009/3300 [01:02<37:00,  1.72s/it, lr: 1.0e-04 loss: 1.492e-01]
anha:  61%|######    | 2009/3300 [01:02<37:00,  1.72s/it, lr: 1.0e-04 loss: 1.492e-01]
anha:  61%|######    | 2010/3300 [01:03<36:59,  1.72s/it, lr: 1.0e-04 loss: 5.491e-02]
anha:  61%|######    | 2010/3300 [01:03<36:59,  1.72s/it, lr: 1.0e-04 loss: 5.491e-02]
anha:  61%|######    | 2011/3300 [01:03<25:00,  1.16s/it, lr: 1.0e-04 loss: 5.491e-02]
anha:  61%|######    | 2011/3300 [01:03<25:00,  1.16s/it, lr: 1.0e-04 loss: 5.491e-02]
anha:  61%|######    | 2011/3300 [01:05<25:00,  1.16s/it, lr: 1.0e-04 loss: 8.827e-02]
anha:  61%|######    | 2011/3300 [01:05<25:00,  1.16s/it, lr: 1.0e-04 loss: 8.827e-02]
anha:  61%|######    | 2012/3300 [01:05<27:52,  1.30s/it, lr: 1.0e-04 loss: 8.827e-02]
anha:  61%|######    | 2012/3300 [01:05<27:52,  1.30s/it, lr: 1.0e-04 loss: 8.827e-02]
anha:  61%|######    | 2012/3300 [01:08<27:52,  1.30s/it, lr: 1.0e-04 loss: 6.507e-02]
anha:  61%|######    | 2012/3300 [01:08<27:52,  1.30s/it, lr: 1.0e-04 loss: 6.507e-02]
anha:  61%|######1   | 2013/3300 [01:08<39:49,  1.86s/it, lr: 1.0e-04 loss: 6.507e-02]
anha:  61%|######1   | 2013/3300 [01:08<39:49,  1.86s/it, lr: 1.0e-04 loss: 6.507e-02]
anha:  61%|######1   | 2013/3300 [01:10<39:49,  1.86s/it, lr: 1.0e-04 loss: 1.079e-01]
anha:  61%|######1   | 2013/3300 [01:10<39:49,  1.86s/it, lr: 1.0e-04 loss: 1.079e-01]
anha:  61%|######1   | 2014/3300 [01:10<39:03,  1.82s/it, lr: 1.0e-04 loss: 1.079e-01]
anha:  61%|######1   | 2014/3300 [01:10<39:03,  1.82s/it, lr: 1.0e-04 loss: 1.079e-01]
anha:  61%|######1   | 2014/3300 [01:11<39:03,  1.82s/it, lr: 1.0e-04 loss: 5.380e-02]
anha:  61%|######1   | 2014/3300 [01:11<39:03,  1.82s/it, lr: 1.0e-04 loss: 5.380e-02]
anha:  61%|######1   | 2015/3300 [01:11<34:08,  1.59s/it, lr: 1.0e-04 loss: 5.380e-02]
anha:  61%|######1   | 2015/3300 [01:11<34:08,  1.59s/it, lr: 1.0e-04 loss: 5.380e-02]
anha:  61%|######1   | 2015/3300 [01:13<34:08,  1.59s/it, lr: 1.0e-04 loss: 6.058e-02]
anha:  61%|######1   | 2015/3300 [01:13<34:08,  1.59s/it, lr: 1.0e-04 loss: 6.058e-02]
anha:  61%|######1   | 2016/3300 [01:13<34:57,  1.63s/it, lr: 1.0e-04 loss: 6.058e-02]
anha:  61%|######1   | 2016/3300 [01:13<34:57,  1.63s/it, lr: 1.0e-04 loss: 6.058e-02]
anha:  61%|######1   | 2016/3300 [01:15<34:57,  1.63s/it, lr: 1.0e-04 loss: 6.482e-02]
anha:  61%|######1   | 2016/3300 [01:15<34:57,  1.63s/it, lr: 1.0e-04 loss: 6.482e-02]
anha:  61%|######1   | 2017/3300 [01:15<36:58,  1.73s/it, lr: 1.0e-04 loss: 6.482e-02]
anha:  61%|######1   | 2017/3300 [01:15<36:58,  1.73s/it, lr: 1.0e-04 loss: 6.482e-02]
anha:  61%|######1   | 2017/3300 [01:17<36:58,  1.73s/it, lr: 1.0e-04 loss: 4.657e-02]
anha:  61%|######1   | 2017/3300 [01:17<36:58,  1.73s/it, lr: 1.0e-04 loss: 4.657e-02]
anha:  61%|######1   | 2018/3300 [01:17<41:49,  1.96s/it, lr: 1.0e-04 loss: 4.657e-02]
anha:  61%|######1   | 2018/3300 [01:17<41:49,  1.96s/it, lr: 1.0e-04 loss: 4.657e-02]
anha:  61%|######1   | 2018/3300 [01:19<41:49,  1.96s/it, lr: 1.0e-04 loss: 7.513e-02]
anha:  61%|######1   | 2018/3300 [01:19<41:49,  1.96s/it, lr: 1.0e-04 loss: 7.513e-02]
anha:  61%|######1   | 2019/3300 [01:19<42:13,  1.98s/it, lr: 1.0e-04 loss: 7.513e-02]
anha:  61%|######1   | 2019/3300 [01:19<42:13,  1.98s/it, lr: 1.0e-04 loss: 7.513e-02]
anha:  61%|######1   | 2019/3300 [01:21<42:13,  1.98s/it, lr: 1.0e-04 loss: 5.601e-02]
anha:  61%|######1   | 2019/3300 [01:21<42:13,  1.98s/it, lr: 1.0e-04 loss: 5.601e-02]
anha:  61%|######1   | 2020/3300 [01:23<42:11,  1.98s/it, lr: 1.0e-04 loss: 5.704e-02]
anha:  61%|######1   | 2020/3300 [01:23<42:11,  1.98s/it, lr: 1.0e-04 loss: 5.704e-02]
anha:  61%|######1   | 2021/3300 [01:23<31:18,  1.47s/it, lr: 1.0e-04 loss: 5.704e-02]
anha:  61%|######1   | 2021/3300 [01:23<31:18,  1.47s/it, lr: 1.0e-04 loss: 5.704e-02]
anha:  61%|######1   | 2021/3300 [01:24<31:18,  1.47s/it, lr: 1.0e-04 loss: 1.110e-01]
anha:  61%|######1   | 2021/3300 [01:24<31:18,  1.47s/it, lr: 1.0e-04 loss: 1.110e-01]
anha:  61%|######1   | 2022/3300 [01:24<28:46,  1.35s/it, lr: 1.0e-04 loss: 1.110e-01]
anha:  61%|######1   | 2022/3300 [01:24<28:46,  1.35s/it, lr: 1.0e-04 loss: 1.110e-01]
anha:  61%|######1   | 2022/3300 [01:26<28:46,  1.35s/it, lr: 1.0e-04 loss: 1.921e-01]
anha:  61%|######1   | 2022/3300 [01:26<28:46,  1.35s/it, lr: 1.0e-04 loss: 1.921e-01]
anha:  61%|######1   | 2023/3300 [01:26<32:04,  1.51s/it, lr: 1.0e-04 loss: 1.921e-01]
anha:  61%|######1   | 2023/3300 [01:26<32:04,  1.51s/it, lr: 1.0e-04 loss: 1.921e-01]
anha:  61%|######1   | 2023/3300 [01:28<32:04,  1.51s/it, lr: 1.0e-04 loss: 5.402e-02]
anha:  61%|######1   | 2023/3300 [01:28<32:04,  1.51s/it, lr: 1.0e-04 loss: 5.402e-02]
anha:  61%|######1   | 2024/3300 [01:28<34:45,  1.63s/it, lr: 1.0e-04 loss: 5.402e-02]
anha:  61%|######1   | 2024/3300 [01:28<34:45,  1.63s/it, lr: 1.0e-04 loss: 5.402e-02]
anha:  61%|######1   | 2024/3300 [01:31<34:45,  1.63s/it, lr: 1.0e-04 loss: 4.931e-02]
anha:  61%|######1   | 2024/3300 [01:31<34:45,  1.63s/it, lr: 1.0e-04 loss: 4.931e-02]
anha:  61%|######1   | 2025/3300 [01:31<45:02,  2.12s/it, lr: 1.0e-04 loss: 4.931e-02]
anha:  61%|######1   | 2025/3300 [01:31<45:02,  2.12s/it, lr: 1.0e-04 loss: 4.931e-02]
anha:  61%|######1   | 2025/3300 [01:33<45:02,  2.12s/it, lr: 1.0e-04 loss: 9.175e-02]
anha:  61%|######1   | 2025/3300 [01:33<45:02,  2.12s/it, lr: 1.0e-04 loss: 9.175e-02]
anha:  61%|######1   | 2026/3300 [01:33<43:59,  2.07s/it, lr: 1.0e-04 loss: 9.175e-02]
anha:  61%|######1   | 2026/3300 [01:33<43:59,  2.07s/it, lr: 1.0e-04 loss: 9.175e-02]
anha:  61%|######1   | 2026/3300 [01:34<43:59,  2.07s/it, lr: 1.0e-04 loss: 7.725e-02]
anha:  61%|######1   | 2026/3300 [01:34<43:59,  2.07s/it, lr: 1.0e-04 loss: 7.725e-02]
anha:  61%|######1   | 2027/3300 [01:34<37:20,  1.76s/it, lr: 1.0e-04 loss: 7.725e-02]
anha:  61%|######1   | 2027/3300 [01:34<37:20,  1.76s/it, lr: 1.0e-04 loss: 7.725e-02]
anha:  61%|######1   | 2027/3300 [01:36<37:20,  1.76s/it, lr: 1.0e-04 loss: 1.664e-01]
anha:  61%|######1   | 2027/3300 [01:36<37:20,  1.76s/it, lr: 1.0e-04 loss: 1.664e-01]
anha:  61%|######1   | 2028/3300 [01:36<38:29,  1.82s/it, lr: 1.0e-04 loss: 1.664e-01]
anha:  61%|######1   | 2028/3300 [01:36<38:29,  1.82s/it, lr: 1.0e-04 loss: 1.664e-01]
anha:  61%|######1   | 2028/3300 [01:38<38:29,  1.82s/it, lr: 1.0e-04 loss: 4.375e-02]
anha:  61%|######1   | 2028/3300 [01:38<38:29,  1.82s/it, lr: 1.0e-04 loss: 4.375e-02]
anha:  61%|######1   | 2029/3300 [01:38<38:37,  1.82s/it, lr: 1.0e-04 loss: 4.375e-02]
anha:  61%|######1   | 2029/3300 [01:38<38:37,  1.82s/it, lr: 1.0e-04 loss: 4.375e-02]
anha:  61%|######1   | 2029/3300 [01:41<38:37,  1.82s/it, lr: 1.0e-04 loss: 1.555e-01]
anha:  61%|######1   | 2029/3300 [01:41<38:37,  1.82s/it, lr: 1.0e-04 loss: 1.555e-01]
anha:  62%|######1   | 2030/3300 [01:45<38:35,  1.82s/it, lr: 1.0e-04 loss: 1.172e-01]
anha:  62%|######1   | 2030/3300 [01:45<38:35,  1.82s/it, lr: 1.0e-04 loss: 1.172e-01]
anha:  62%|######1   | 2031/3300 [01:45<36:51,  1.74s/it, lr: 1.0e-04 loss: 1.172e-01]
anha:  62%|######1   | 2031/3300 [01:45<36:51,  1.74s/it, lr: 1.0e-04 loss: 1.172e-01]
anha:  62%|######1   | 2031/3300 [01:46<36:51,  1.74s/it, lr: 1.0e-04 loss: 5.663e-02]
anha:  62%|######1   | 2031/3300 [01:46<36:51,  1.74s/it, lr: 1.0e-04 loss: 5.663e-02]
anha:  62%|######1   | 2032/3300 [01:46<33:00,  1.56s/it, lr: 1.0e-04 loss: 5.663e-02]
anha:  62%|######1   | 2032/3300 [01:46<33:00,  1.56s/it, lr: 1.0e-04 loss: 5.663e-02]
anha:  62%|######1   | 2032/3300 [01:49<33:00,  1.56s/it, lr: 1.0e-04 loss: 5.820e-02]
anha:  62%|######1   | 2032/3300 [01:49<33:00,  1.56s/it, lr: 1.0e-04 loss: 5.820e-02]
anha:  62%|######1   | 2033/3300 [01:49<43:19,  2.05s/it, lr: 1.0e-04 loss: 5.820e-02]
anha:  62%|######1   | 2033/3300 [01:49<43:19,  2.05s/it, lr: 1.0e-04 loss: 5.820e-02]
anha:  62%|######1   | 2033/3300 [01:50<43:19,  2.05s/it, lr: 1.0e-04 loss: 4.414e-02]
anha:  62%|######1   | 2033/3300 [01:50<43:19,  2.05s/it, lr: 1.0e-04 loss: 4.414e-02]
anha:  62%|######1   | 2034/3300 [01:50<37:17,  1.77s/it, lr: 1.0e-04 loss: 4.414e-02]
anha:  62%|######1   | 2034/3300 [01:50<37:17,  1.77s/it, lr: 1.0e-04 loss: 4.414e-02]
anha:  62%|######1   | 2034/3300 [01:52<37:17,  1.77s/it, lr: 1.0e-04 loss: 2.608e-02]
anha:  62%|######1   | 2034/3300 [01:52<37:17,  1.77s/it, lr: 1.0e-04 loss: 2.608e-02]
anha:  62%|######1   | 2035/3300 [01:52<38:21,  1.82s/it, lr: 1.0e-04 loss: 2.608e-02]
anha:  62%|######1   | 2035/3300 [01:52<38:21,  1.82s/it, lr: 1.0e-04 loss: 2.608e-02]
anha:  62%|######1   | 2035/3300 [01:54<38:21,  1.82s/it, lr: 1.0e-04 loss: 4.401e-02]
anha:  62%|######1   | 2035/3300 [01:54<38:21,  1.82s/it, lr: 1.0e-04 loss: 4.401e-02]
anha:  62%|######1   | 2036/3300 [01:54<37:43,  1.79s/it, lr: 1.0e-04 loss: 4.401e-02]
anha:  62%|######1   | 2036/3300 [01:54<37:43,  1.79s/it, lr: 1.0e-04 loss: 4.401e-02]
anha:  62%|######1   | 2036/3300 [01:57<37:43,  1.79s/it, lr: 1.0e-04 loss: 7.608e-02]
anha:  62%|######1   | 2036/3300 [01:57<37:43,  1.79s/it, lr: 1.0e-04 loss: 7.608e-02]
anha:  62%|######1   | 2037/3300 [01:57<46:29,  2.21s/it, lr: 1.0e-04 loss: 7.608e-02]
anha:  62%|######1   | 2037/3300 [01:57<46:29,  2.21s/it, lr: 1.0e-04 loss: 7.608e-02]
anha:  62%|######1   | 2037/3300 [01:58<46:29,  2.21s/it, lr: 1.0e-04 loss: 9.276e-02]
anha:  62%|######1   | 2037/3300 [01:58<46:29,  2.21s/it, lr: 1.0e-04 loss: 9.276e-02]
anha:  62%|######1   | 2038/3300 [01:58<39:43,  1.89s/it, lr: 1.0e-04 loss: 9.276e-02]
anha:  62%|######1   | 2038/3300 [01:58<39:43,  1.89s/it, lr: 1.0e-04 loss: 9.276e-02]
anha:  62%|######1   | 2038/3300 [02:02<39:43,  1.89s/it, lr: 1.0e-04 loss: 3.784e-02]
anha:  62%|######1   | 2038/3300 [02:02<39:43,  1.89s/it, lr: 1.0e-04 loss: 3.784e-02]
anha:  62%|######1   | 2039/3300 [02:02<49:33,  2.36s/it, lr: 1.0e-04 loss: 3.784e-02]
anha:  62%|######1   | 2039/3300 [02:02<49:33,  2.36s/it, lr: 1.0e-04 loss: 3.784e-02]
anha:  62%|######1   | 2039/3300 [02:05<49:33,  2.36s/it, lr: 1.0e-04 loss: 6.418e-02]
anha:  62%|######1   | 2039/3300 [02:05<49:33,  2.36s/it, lr: 1.0e-04 loss: 6.418e-02]
anha:  62%|######1   | 2040/3300 [02:06<49:30,  2.36s/it, lr: 1.0e-04 loss: 2.555e-01]
anha:  62%|######1   | 2040/3300 [02:06<49:30,  2.36s/it, lr: 1.0e-04 loss: 2.555e-01]
anha:  62%|######1   | 2041/3300 [02:06<31:39,  1.51s/it, lr: 1.0e-04 loss: 2.555e-01]
anha:  62%|######1   | 2041/3300 [02:06<31:39,  1.51s/it, lr: 1.0e-04 loss: 2.555e-01]
anha:  62%|######1   | 2041/3300 [02:08<31:39,  1.51s/it, lr: 1.0e-04 loss: 3.889e-02]
anha:  62%|######1   | 2041/3300 [02:08<31:39,  1.51s/it, lr: 1.0e-04 loss: 3.889e-02]
anha:  62%|######1   | 2042/3300 [02:08<34:03,  1.62s/it, lr: 1.0e-04 loss: 3.889e-02]
anha:  62%|######1   | 2042/3300 [02:08<34:03,  1.62s/it, lr: 1.0e-04 loss: 3.889e-02]
anha:  62%|######1   | 2042/3300 [02:11<34:03,  1.62s/it, lr: 1.0e-04 loss: 1.151e-01]
anha:  62%|######1   | 2042/3300 [02:11<34:03,  1.62s/it, lr: 1.0e-04 loss: 1.151e-01]
anha:  62%|######1   | 2043/3300 [02:11<44:16,  2.11s/it, lr: 1.0e-04 loss: 1.151e-01]
anha:  62%|######1   | 2043/3300 [02:11<44:16,  2.11s/it, lr: 1.0e-04 loss: 1.151e-01]
anha:  62%|######1   | 2043/3300 [02:15<44:16,  2.11s/it, lr: 1.0e-04 loss: 4.233e-02]
anha:  62%|######1   | 2043/3300 [02:15<44:16,  2.11s/it, lr: 1.0e-04 loss: 4.233e-02]
anha:  62%|######1   | 2044/3300 [02:15<50:09,  2.40s/it, lr: 1.0e-04 loss: 4.233e-02]
anha:  62%|######1   | 2044/3300 [02:15<50:09,  2.40s/it, lr: 1.0e-04 loss: 4.233e-02]
anha:  62%|######1   | 2044/3300 [02:18<50:09,  2.40s/it, lr: 1.0e-04 loss: 3.386e-02]
anha:  62%|######1   | 2044/3300 [02:18<50:09,  2.40s/it, lr: 1.0e-04 loss: 3.386e-02]
anha:  62%|######1   | 2045/3300 [02:18<55:48,  2.67s/it, lr: 1.0e-04 loss: 3.386e-02]
anha:  62%|######1   | 2045/3300 [02:18<55:48,  2.67s/it, lr: 1.0e-04 loss: 3.386e-02]
anha:  62%|######1   | 2045/3300 [02:21<55:48,  2.67s/it, lr: 1.0e-04 loss: 6.225e-02]
anha:  62%|######1   | 2045/3300 [02:21<55:48,  2.67s/it, lr: 1.0e-04 loss: 6.225e-02]
anha:  62%|######2   | 2046/3300 [02:21<59:25,  2.84s/it, lr: 1.0e-04 loss: 6.225e-02]
anha:  62%|######2   | 2046/3300 [02:21<59:25,  2.84s/it, lr: 1.0e-04 loss: 6.225e-02]
anha:  62%|######2   | 2046/3300 [02:25<59:25,  2.84s/it, lr: 1.0e-04 loss: 1.100e-01]
anha:  62%|######2   | 2046/3300 [02:25<59:25,  2.84s/it, lr: 1.0e-04 loss: 1.100e-01]
anha:  62%|######2   | 2047/3300 [02:25<1:03:36,  3.05s/it, lr: 1.0e-04 loss: 1.100e-01]
anha:  62%|######2   | 2047/3300 [02:25<1:03:36,  3.05s/it, lr: 1.0e-04 loss: 1.100e-01]
anha:  62%|######2   | 2047/3300 [02:27<1:03:36,  3.05s/it, lr: 1.0e-04 loss: 3.708e-02]
anha:  62%|######2   | 2047/3300 [02:27<1:03:36,  3.05s/it, lr: 1.0e-04 loss: 3.708e-02]
anha:  62%|######2   | 2048/3300 [02:27<57:10,  2.74s/it, lr: 1.0e-04 loss: 3.708e-02]  
anha:  62%|######2   | 2048/3300 [02:27<57:10,  2.74s/it, lr: 1.0e-04 loss: 3.708e-02]  
anha:  62%|######2   | 2048/3300 [02:29<57:10,  2.74s/it, lr: 1.0e-04 loss: 8.988e-02]
anha:  62%|######2   | 2048/3300 [02:29<57:10,  2.74s/it, lr: 1.0e-04 loss: 8.988e-02]
anha:  62%|######2   | 2049/3300 [02:29<52:17,  2.51s/it, lr: 1.0e-04 loss: 8.988e-02]
anha:  62%|######2   | 2049/3300 [02:29<52:17,  2.51s/it, lr: 1.0e-04 loss: 8.988e-02]
anha:  62%|######2   | 2049/3300 [02:31<52:17,  2.51s/it, lr: 1.0e-04 loss: 8.203e-02]
anha:  62%|######2   | 2049/3300 [02:31<52:17,  2.51s/it, lr: 1.0e-04 loss: 8.203e-02]
anha:  62%|######2   | 2050/3300 [02:34<52:15,  2.51s/it, lr: 1.0e-04 loss: 4.882e-02]
anha:  62%|######2   | 2050/3300 [02:34<52:15,  2.51s/it, lr: 1.0e-04 loss: 4.882e-02]
anha:  62%|######2   | 2051/3300 [02:34<44:22,  2.13s/it, lr: 1.0e-04 loss: 4.882e-02]
anha:  62%|######2   | 2051/3300 [02:34<44:22,  2.13s/it, lr: 1.0e-04 loss: 4.882e-02]
anha:  62%|######2   | 2051/3300 [02:35<44:22,  2.13s/it, lr: 1.0e-04 loss: 1.561e-01]
anha:  62%|######2   | 2051/3300 [02:35<44:22,  2.13s/it, lr: 1.0e-04 loss: 1.561e-01]
anha:  62%|######2   | 2052/3300 [02:35<38:34,  1.85s/it, lr: 1.0e-04 loss: 1.561e-01]
anha:  62%|######2   | 2052/3300 [02:35<38:34,  1.85s/it, lr: 1.0e-04 loss: 1.561e-01]
anha:  62%|######2   | 2052/3300 [02:36<38:34,  1.85s/it, lr: 1.0e-04 loss: 7.802e-02]
anha:  62%|######2   | 2052/3300 [02:36<38:34,  1.85s/it, lr: 1.0e-04 loss: 7.802e-02]
anha:  62%|######2   | 2053/3300 [02:36<33:52,  1.63s/it, lr: 1.0e-04 loss: 7.802e-02]
anha:  62%|######2   | 2053/3300 [02:36<33:52,  1.63s/it, lr: 1.0e-04 loss: 7.802e-02]
anha:  62%|######2   | 2053/3300 [02:37<33:52,  1.63s/it, lr: 1.0e-04 loss: 1.033e-01]
anha:  62%|######2   | 2053/3300 [02:37<33:52,  1.63s/it, lr: 1.0e-04 loss: 1.033e-01]
anha:  62%|######2   | 2054/3300 [02:37<30:15,  1.46s/it, lr: 1.0e-04 loss: 1.033e-01]
anha:  62%|######2   | 2054/3300 [02:37<30:15,  1.46s/it, lr: 1.0e-04 loss: 1.033e-01]
anha:  62%|######2   | 2054/3300 [02:38<30:15,  1.46s/it, lr: 1.0e-04 loss: 5.014e-02]
anha:  62%|######2   | 2054/3300 [02:38<30:15,  1.46s/it, lr: 1.0e-04 loss: 5.014e-02]
anha:  62%|######2   | 2055/3300 [02:38<27:38,  1.33s/it, lr: 1.0e-04 loss: 5.014e-02]
anha:  62%|######2   | 2055/3300 [02:38<27:38,  1.33s/it, lr: 1.0e-04 loss: 5.014e-02]
anha:  62%|######2   | 2055/3300 [02:40<27:38,  1.33s/it, lr: 1.0e-04 loss: 4.307e-02]
anha:  62%|######2   | 2055/3300 [02:40<27:38,  1.33s/it, lr: 1.0e-04 loss: 4.307e-02]
anha:  62%|######2   | 2056/3300 [02:40<31:17,  1.51s/it, lr: 1.0e-04 loss: 4.307e-02]
anha:  62%|######2   | 2056/3300 [02:40<31:17,  1.51s/it, lr: 1.0e-04 loss: 4.307e-02]
anha:  62%|######2   | 2056/3300 [02:42<31:17,  1.51s/it, lr: 1.0e-04 loss: 5.031e-02]
anha:  62%|######2   | 2056/3300 [02:42<31:17,  1.51s/it, lr: 1.0e-04 loss: 5.031e-02]
anha:  62%|######2   | 2057/3300 [02:42<34:27,  1.66s/it, lr: 1.0e-04 loss: 5.031e-02]
anha:  62%|######2   | 2057/3300 [02:42<34:27,  1.66s/it, lr: 1.0e-04 loss: 5.031e-02]
anha:  62%|######2   | 2057/3300 [02:44<34:27,  1.66s/it, lr: 1.0e-04 loss: 8.671e-02]
anha:  62%|######2   | 2057/3300 [02:44<34:27,  1.66s/it, lr: 1.0e-04 loss: 8.671e-02]
anha:  62%|######2   | 2058/3300 [02:44<36:26,  1.76s/it, lr: 1.0e-04 loss: 8.671e-02]
anha:  62%|######2   | 2058/3300 [02:44<36:26,  1.76s/it, lr: 1.0e-04 loss: 8.671e-02]
anha:  62%|######2   | 2058/3300 [02:47<36:26,  1.76s/it, lr: 1.0e-04 loss: 2.919e-02]
anha:  62%|######2   | 2058/3300 [02:47<36:26,  1.76s/it, lr: 1.0e-04 loss: 2.919e-02]
anha:  62%|######2   | 2059/3300 [02:47<46:21,  2.24s/it, lr: 1.0e-04 loss: 2.919e-02]
anha:  62%|######2   | 2059/3300 [02:47<46:21,  2.24s/it, lr: 1.0e-04 loss: 2.919e-02]
anha:  62%|######2   | 2059/3300 [02:51<46:21,  2.24s/it, lr: 1.0e-04 loss: 1.488e-01]
anha:  62%|######2   | 2059/3300 [02:51<46:21,  2.24s/it, lr: 1.0e-04 loss: 1.488e-01]
anha:  62%|######2   | 2060/3300 [02:54<46:19,  2.24s/it, lr: 1.0e-04 loss: 6.165e-02]
anha:  62%|######2   | 2060/3300 [02:54<46:19,  2.24s/it, lr: 1.0e-04 loss: 6.165e-02]
anha:  62%|######2   | 2061/3300 [02:54<41:31,  2.01s/it, lr: 1.0e-04 loss: 6.165e-02]
anha:  62%|######2   | 2061/3300 [02:54<41:31,  2.01s/it, lr: 1.0e-04 loss: 6.165e-02]
anha:  62%|######2   | 2061/3300 [02:55<41:31,  2.01s/it, lr: 1.0e-04 loss: 5.392e-01]
anha:  62%|######2   | 2061/3300 [02:55<41:31,  2.01s/it, lr: 1.0e-04 loss: 5.392e-01]
anha:  62%|######2   | 2062/3300 [02:55<36:22,  1.76s/it, lr: 1.0e-04 loss: 5.392e-01]
anha:  62%|######2   | 2062/3300 [02:55<36:22,  1.76s/it, lr: 1.0e-04 loss: 5.392e-01]
anha:  62%|######2   | 2062/3300 [02:56<36:22,  1.76s/it, lr: 1.0e-04 loss: 5.726e-02]
anha:  62%|######2   | 2062/3300 [02:56<36:22,  1.76s/it, lr: 1.0e-04 loss: 5.726e-02]
anha:  63%|######2   | 2063/3300 [02:56<32:16,  1.57s/it, lr: 1.0e-04 loss: 5.726e-02]
anha:  63%|######2   | 2063/3300 [02:56<32:16,  1.57s/it, lr: 1.0e-04 loss: 5.726e-02]
anha:  63%|######2   | 2063/3300 [03:00<32:16,  1.57s/it, lr: 1.0e-04 loss: 4.042e-02]
anha:  63%|######2   | 2063/3300 [03:00<32:16,  1.57s/it, lr: 1.0e-04 loss: 4.042e-02]
anha:  63%|######2   | 2064/3300 [03:00<42:19,  2.05s/it, lr: 1.0e-04 loss: 4.042e-02]
anha:  63%|######2   | 2064/3300 [03:00<42:19,  2.05s/it, lr: 1.0e-04 loss: 4.042e-02]
anha:  63%|######2   | 2064/3300 [03:02<42:19,  2.05s/it, lr: 1.0e-04 loss: 7.902e-02]
anha:  63%|######2   | 2064/3300 [03:02<42:19,  2.05s/it, lr: 1.0e-04 loss: 7.902e-02]
anha:  63%|######2   | 2065/3300 [03:02<41:48,  2.03s/it, lr: 1.0e-04 loss: 7.902e-02]
anha:  63%|######2   | 2065/3300 [03:02<41:48,  2.03s/it, lr: 1.0e-04 loss: 7.902e-02]
anha:  63%|######2   | 2065/3300 [03:05<41:48,  2.03s/it, lr: 1.0e-04 loss: 9.723e-02]
anha:  63%|######2   | 2065/3300 [03:05<41:48,  2.03s/it, lr: 1.0e-04 loss: 9.723e-02]
anha:  63%|######2   | 2066/3300 [03:05<49:54,  2.43s/it, lr: 1.0e-04 loss: 9.723e-02]
anha:  63%|######2   | 2066/3300 [03:05<49:54,  2.43s/it, lr: 1.0e-04 loss: 9.723e-02]
anha:  63%|######2   | 2066/3300 [03:08<49:54,  2.43s/it, lr: 1.0e-04 loss: 1.079e-01]
anha:  63%|######2   | 2066/3300 [03:08<49:54,  2.43s/it, lr: 1.0e-04 loss: 1.079e-01]
anha:  63%|######2   | 2067/3300 [03:08<54:41,  2.66s/it, lr: 1.0e-04 loss: 1.079e-01]
anha:  63%|######2   | 2067/3300 [03:08<54:41,  2.66s/it, lr: 1.0e-04 loss: 1.079e-01]
anha:  63%|######2   | 2067/3300 [03:12<54:41,  2.66s/it, lr: 1.0e-04 loss: 4.832e-02]
anha:  63%|######2   | 2067/3300 [03:12<54:41,  2.66s/it, lr: 1.0e-04 loss: 4.832e-02]
anha:  63%|######2   | 2068/3300 [03:12<59:33,  2.90s/it, lr: 1.0e-04 loss: 4.832e-02]
anha:  63%|######2   | 2068/3300 [03:12<59:33,  2.90s/it, lr: 1.0e-04 loss: 4.832e-02]
anha:  63%|######2   | 2068/3300 [03:14<59:33,  2.90s/it, lr: 1.0e-04 loss: 1.035e-01]
anha:  63%|######2   | 2068/3300 [03:14<59:33,  2.90s/it, lr: 1.0e-04 loss: 1.035e-01]
anha:  63%|######2   | 2069/3300 [03:14<53:51,  2.63s/it, lr: 1.0e-04 loss: 1.035e-01]
anha:  63%|######2   | 2069/3300 [03:14<53:51,  2.63s/it, lr: 1.0e-04 loss: 1.035e-01]
anha:  63%|######2   | 2069/3300 [03:16<53:51,  2.63s/it, lr: 1.0e-04 loss: 7.684e-02]
anha:  63%|######2   | 2069/3300 [03:16<53:51,  2.63s/it, lr: 1.0e-04 loss: 7.684e-02]
anha:  63%|######2   | 2070/3300 [03:17<53:49,  2.63s/it, lr: 1.0e-04 loss: 1.040e-01]
anha:  63%|######2   | 2070/3300 [03:17<53:49,  2.63s/it, lr: 1.0e-04 loss: 1.040e-01]
anha:  63%|######2   | 2071/3300 [03:17<37:17,  1.82s/it, lr: 1.0e-04 loss: 1.040e-01]
anha:  63%|######2   | 2071/3300 [03:17<37:17,  1.82s/it, lr: 1.0e-04 loss: 1.040e-01]
anha:  63%|######2   | 2071/3300 [03:21<37:17,  1.82s/it, lr: 1.0e-04 loss: 2.447e-01]
anha:  63%|######2   | 2071/3300 [03:21<37:17,  1.82s/it, lr: 1.0e-04 loss: 2.447e-01]
anha:  63%|######2   | 2072/3300 [03:21<44:39,  2.18s/it, lr: 1.0e-04 loss: 2.447e-01]
anha:  63%|######2   | 2072/3300 [03:21<44:39,  2.18s/it, lr: 1.0e-04 loss: 2.447e-01]
anha:  63%|######2   | 2072/3300 [03:22<44:39,  2.18s/it, lr: 1.0e-04 loss: 8.167e-02]
anha:  63%|######2   | 2072/3300 [03:22<44:39,  2.18s/it, lr: 1.0e-04 loss: 8.167e-02]
anha:  63%|######2   | 2073/3300 [03:22<42:09,  2.06s/it, lr: 1.0e-04 loss: 8.167e-02]
anha:  63%|######2   | 2073/3300 [03:22<42:09,  2.06s/it, lr: 1.0e-04 loss: 8.167e-02]
anha:  63%|######2   | 2073/3300 [03:24<42:09,  2.06s/it, lr: 1.0e-04 loss: 4.939e-02]
anha:  63%|######2   | 2073/3300 [03:24<42:09,  2.06s/it, lr: 1.0e-04 loss: 4.939e-02]
anha:  63%|######2   | 2074/3300 [03:24<41:39,  2.04s/it, lr: 1.0e-04 loss: 4.939e-02]
anha:  63%|######2   | 2074/3300 [03:24<41:39,  2.04s/it, lr: 1.0e-04 loss: 4.939e-02]
anha:  63%|######2   | 2074/3300 [03:25<41:39,  2.04s/it, lr: 1.0e-04 loss: 5.892e-02]
anha:  63%|######2   | 2074/3300 [03:25<41:39,  2.04s/it, lr: 1.0e-04 loss: 5.892e-02]
anha:  63%|######2   | 2075/3300 [03:25<35:45,  1.75s/it, lr: 1.0e-04 loss: 5.892e-02]
anha:  63%|######2   | 2075/3300 [03:25<35:45,  1.75s/it, lr: 1.0e-04 loss: 5.892e-02]
anha:  63%|######2   | 2075/3300 [03:26<35:45,  1.75s/it, lr: 1.0e-04 loss: 1.595e-01]
anha:  63%|######2   | 2075/3300 [03:26<35:45,  1.75s/it, lr: 1.0e-04 loss: 1.595e-01]
anha:  63%|######2   | 2076/3300 [03:26<32:01,  1.57s/it, lr: 1.0e-04 loss: 1.595e-01]
anha:  63%|######2   | 2076/3300 [03:26<32:01,  1.57s/it, lr: 1.0e-04 loss: 1.595e-01]
anha:  63%|######2   | 2076/3300 [03:30<32:01,  1.57s/it, lr: 1.0e-04 loss: 2.579e-02]
anha:  63%|######2   | 2076/3300 [03:30<32:01,  1.57s/it, lr: 1.0e-04 loss: 2.579e-02]
anha:  63%|######2   | 2077/3300 [03:30<42:37,  2.09s/it, lr: 1.0e-04 loss: 2.579e-02]
anha:  63%|######2   | 2077/3300 [03:30<42:37,  2.09s/it, lr: 1.0e-04 loss: 2.579e-02]
anha:  63%|######2   | 2077/3300 [03:31<42:37,  2.09s/it, lr: 1.0e-04 loss: 7.893e-02]
anha:  63%|######2   | 2077/3300 [03:31<42:37,  2.09s/it, lr: 1.0e-04 loss: 7.893e-02]
anha:  63%|######2   | 2078/3300 [03:31<36:10,  1.78s/it, lr: 1.0e-04 loss: 7.893e-02]
anha:  63%|######2   | 2078/3300 [03:31<36:10,  1.78s/it, lr: 1.0e-04 loss: 7.893e-02]
anha:  63%|######2   | 2078/3300 [03:32<36:10,  1.78s/it, lr: 1.0e-04 loss: 6.476e-02]
anha:  63%|######2   | 2078/3300 [03:32<36:10,  1.78s/it, lr: 1.0e-04 loss: 6.476e-02]
anha:  63%|######3   | 2079/3300 [03:32<31:27,  1.55s/it, lr: 1.0e-04 loss: 6.476e-02]
anha:  63%|######3   | 2079/3300 [03:32<31:27,  1.55s/it, lr: 1.0e-04 loss: 6.476e-02]
anha:  63%|######3   | 2079/3300 [03:33<31:27,  1.55s/it, lr: 1.0e-04 loss: 5.375e-02]
anha:  63%|######3   | 2079/3300 [03:33<31:27,  1.55s/it, lr: 1.0e-04 loss: 5.375e-02]
anha:  63%|######3   | 2080/3300 [03:36<31:25,  1.55s/it, lr: 1.0e-04 loss: 1.323e-01]
anha:  63%|######3   | 2080/3300 [03:36<31:25,  1.55s/it, lr: 1.0e-04 loss: 1.323e-01]
anha:  63%|######3   | 2081/3300 [03:36<32:04,  1.58s/it, lr: 1.0e-04 loss: 1.323e-01]
anha:  63%|######3   | 2081/3300 [03:36<32:04,  1.58s/it, lr: 1.0e-04 loss: 1.323e-01]
anha:  63%|######3   | 2081/3300 [03:37<32:04,  1.58s/it, lr: 1.0e-04 loss: 3.541e-01]
anha:  63%|######3   | 2081/3300 [03:37<32:04,  1.58s/it, lr: 1.0e-04 loss: 3.541e-01]
anha:  63%|######3   | 2082/3300 [03:37<29:10,  1.44s/it, lr: 1.0e-04 loss: 3.541e-01]
anha:  63%|######3   | 2082/3300 [03:37<29:10,  1.44s/it, lr: 1.0e-04 loss: 3.541e-01]
anha:  63%|######3   | 2082/3300 [03:41<29:10,  1.44s/it, lr: 1.0e-04 loss: 2.042e-01]
anha:  63%|######3   | 2082/3300 [03:41<29:10,  1.44s/it, lr: 1.0e-04 loss: 2.042e-01]
anha:  63%|######3   | 2083/3300 [03:41<39:56,  1.97s/it, lr: 1.0e-04 loss: 2.042e-01]
anha:  63%|######3   | 2083/3300 [03:41<39:56,  1.97s/it, lr: 1.0e-04 loss: 2.042e-01]
anha:  63%|######3   | 2083/3300 [03:44<39:56,  1.97s/it, lr: 1.0e-04 loss: 1.156e-01]
anha:  63%|######3   | 2083/3300 [03:44<39:56,  1.97s/it, lr: 1.0e-04 loss: 1.156e-01]
anha:  63%|######3   | 2084/3300 [03:44<47:59,  2.37s/it, lr: 1.0e-04 loss: 1.156e-01]
anha:  63%|######3   | 2084/3300 [03:44<47:59,  2.37s/it, lr: 1.0e-04 loss: 1.156e-01]
anha:  63%|######3   | 2084/3300 [03:45<47:59,  2.37s/it, lr: 1.0e-04 loss: 7.539e-02]
anha:  63%|######3   | 2084/3300 [03:45<47:59,  2.37s/it, lr: 1.0e-04 loss: 7.539e-02]
anha:  63%|######3   | 2085/3300 [03:45<40:58,  2.02s/it, lr: 1.0e-04 loss: 7.539e-02]
anha:  63%|######3   | 2085/3300 [03:45<40:58,  2.02s/it, lr: 1.0e-04 loss: 7.539e-02]
anha:  63%|######3   | 2085/3300 [03:49<40:58,  2.02s/it, lr: 1.0e-04 loss: 5.724e-02]
anha:  63%|######3   | 2085/3300 [03:49<40:58,  2.02s/it, lr: 1.0e-04 loss: 5.724e-02]
anha:  63%|######3   | 2086/3300 [03:49<49:21,  2.44s/it, lr: 1.0e-04 loss: 5.724e-02]
anha:  63%|######3   | 2086/3300 [03:49<49:21,  2.44s/it, lr: 1.0e-04 loss: 5.724e-02]
anha:  63%|######3   | 2086/3300 [03:50<49:21,  2.44s/it, lr: 1.0e-04 loss: 3.757e-02]
anha:  63%|######3   | 2086/3300 [03:50<49:21,  2.44s/it, lr: 1.0e-04 loss: 3.757e-02]
anha:  63%|######3   | 2087/3300 [03:50<45:03,  2.23s/it, lr: 1.0e-04 loss: 3.757e-02]
anha:  63%|######3   | 2087/3300 [03:50<45:03,  2.23s/it, lr: 1.0e-04 loss: 3.757e-02]
anha:  63%|######3   | 2087/3300 [03:51<45:03,  2.23s/it, lr: 1.0e-04 loss: 8.623e-02]
anha:  63%|######3   | 2087/3300 [03:51<45:03,  2.23s/it, lr: 1.0e-04 loss: 8.623e-02]
anha:  63%|######3   | 2088/3300 [03:51<37:47,  1.87s/it, lr: 1.0e-04 loss: 8.623e-02]
anha:  63%|######3   | 2088/3300 [03:51<37:47,  1.87s/it, lr: 1.0e-04 loss: 8.623e-02]
anha:  63%|######3   | 2088/3300 [03:52<37:47,  1.87s/it, lr: 1.0e-04 loss: 3.673e-02]
anha:  63%|######3   | 2088/3300 [03:52<37:47,  1.87s/it, lr: 1.0e-04 loss: 3.673e-02]
anha:  63%|######3   | 2089/3300 [03:52<32:33,  1.61s/it, lr: 1.0e-04 loss: 3.673e-02]
anha:  63%|######3   | 2089/3300 [03:52<32:33,  1.61s/it, lr: 1.0e-04 loss: 3.673e-02]
anha:  63%|######3   | 2089/3300 [03:53<32:33,  1.61s/it, lr: 1.0e-04 loss: 6.429e-02]
anha:  63%|######3   | 2089/3300 [03:53<32:33,  1.61s/it, lr: 1.0e-04 loss: 6.429e-02]
anha:  63%|######3   | 2090/3300 [03:54<32:32,  1.61s/it, lr: 1.0e-04 loss: 4.148e-02]
anha:  63%|######3   | 2090/3300 [03:54<32:32,  1.61s/it, lr: 1.0e-04 loss: 4.148e-02]
anha:  63%|######3   | 2091/3300 [03:54<22:17,  1.11s/it, lr: 1.0e-04 loss: 4.148e-02]
anha:  63%|######3   | 2091/3300 [03:54<22:17,  1.11s/it, lr: 1.0e-04 loss: 4.148e-02]
anha:  63%|######3   | 2091/3300 [03:56<22:17,  1.11s/it, lr: 1.0e-04 loss: 2.205e-01]
anha:  63%|######3   | 2091/3300 [03:56<22:17,  1.11s/it, lr: 1.0e-04 loss: 2.205e-01]
anha:  63%|######3   | 2092/3300 [03:56<25:18,  1.26s/it, lr: 1.0e-04 loss: 2.205e-01]
anha:  63%|######3   | 2092/3300 [03:56<25:18,  1.26s/it, lr: 1.0e-04 loss: 2.205e-01]
anha:  63%|######3   | 2092/3300 [03:57<25:18,  1.26s/it, lr: 1.0e-04 loss: 5.755e-02]
anha:  63%|######3   | 2092/3300 [03:57<25:18,  1.26s/it, lr: 1.0e-04 loss: 5.755e-02]
anha:  63%|######3   | 2093/3300 [03:57<23:58,  1.19s/it, lr: 1.0e-04 loss: 5.755e-02]
anha:  63%|######3   | 2093/3300 [03:57<23:58,  1.19s/it, lr: 1.0e-04 loss: 5.755e-02]
anha:  63%|######3   | 2093/3300 [03:59<23:58,  1.19s/it, lr: 1.0e-04 loss: 5.834e-02]
anha:  63%|######3   | 2093/3300 [03:59<23:58,  1.19s/it, lr: 1.0e-04 loss: 5.834e-02]
anha:  63%|######3   | 2094/3300 [03:59<28:30,  1.42s/it, lr: 1.0e-04 loss: 5.834e-02]
anha:  63%|######3   | 2094/3300 [03:59<28:30,  1.42s/it, lr: 1.0e-04 loss: 5.834e-02]
anha:  63%|######3   | 2094/3300 [04:00<28:30,  1.42s/it, lr: 1.0e-04 loss: 8.205e-02]
anha:  63%|######3   | 2094/3300 [04:00<28:30,  1.42s/it, lr: 1.0e-04 loss: 8.205e-02]
anha:  63%|######3   | 2095/3300 [04:00<26:04,  1.30s/it, lr: 1.0e-04 loss: 8.205e-02]
anha:  63%|######3   | 2095/3300 [04:00<26:04,  1.30s/it, lr: 1.0e-04 loss: 8.205e-02]
anha:  63%|######3   | 2095/3300 [04:03<26:04,  1.30s/it, lr: 1.0e-04 loss: 3.419e-01]
anha:  63%|######3   | 2095/3300 [04:03<26:04,  1.30s/it, lr: 1.0e-04 loss: 3.419e-01]
anha:  64%|######3   | 2096/3300 [04:03<37:26,  1.87s/it, lr: 1.0e-04 loss: 3.419e-01]
anha:  64%|######3   | 2096/3300 [04:03<37:26,  1.87s/it, lr: 1.0e-04 loss: 3.419e-01]
anha:  64%|######3   | 2096/3300 [04:07<37:26,  1.87s/it, lr: 1.0e-04 loss: 3.447e-02]
anha:  64%|######3   | 2096/3300 [04:07<37:26,  1.87s/it, lr: 1.0e-04 loss: 3.447e-02]
anha:  64%|######3   | 2097/3300 [04:07<46:53,  2.34s/it, lr: 1.0e-04 loss: 3.447e-02]
anha:  64%|######3   | 2097/3300 [04:07<46:53,  2.34s/it, lr: 1.0e-04 loss: 3.447e-02]
anha:  64%|######3   | 2097/3300 [04:10<46:53,  2.34s/it, lr: 1.0e-04 loss: 3.678e-02]
anha:  64%|######3   | 2097/3300 [04:10<46:53,  2.34s/it, lr: 1.0e-04 loss: 3.678e-02]
anha:  64%|######3   | 2098/3300 [04:10<51:39,  2.58s/it, lr: 1.0e-04 loss: 3.678e-02]
anha:  64%|######3   | 2098/3300 [04:10<51:39,  2.58s/it, lr: 1.0e-04 loss: 3.678e-02]
anha:  64%|######3   | 2098/3300 [04:12<51:39,  2.58s/it, lr: 1.0e-04 loss: 8.068e-02]
anha:  64%|######3   | 2098/3300 [04:12<51:39,  2.58s/it, lr: 1.0e-04 loss: 8.068e-02]
anha:  64%|######3   | 2099/3300 [04:12<46:37,  2.33s/it, lr: 1.0e-04 loss: 8.068e-02]
anha:  64%|######3   | 2099/3300 [04:12<46:37,  2.33s/it, lr: 1.0e-04 loss: 8.068e-02]
anha:  64%|######3   | 2099/3300 [04:14<46:37,  2.33s/it, lr: 1.0e-04 loss: 8.434e-02]
anha:  64%|######3   | 2099/3300 [04:14<46:37,  2.33s/it, lr: 1.0e-04 loss: 8.434e-02]
anha:  64%|######3   | 2100/3300 [04:16<46:34,  2.33s/it, lr: 1.0e-04 loss: 4.634e-02]
anha:  64%|######3   | 2100/3300 [04:16<46:34,  2.33s/it, lr: 1.0e-04 loss: 4.634e-02]
anha:  64%|######3   | 2101/3300 [04:16<34:18,  1.72s/it, lr: 1.0e-04 loss: 4.634e-02]
anha:  64%|######3   | 2101/3300 [04:16<34:18,  1.72s/it, lr: 1.0e-04 loss: 4.634e-02]
anha:  64%|######3   | 2101/3300 [04:17<34:18,  1.72s/it, lr: 1.0e-04 loss: 1.314e-01]
anha:  64%|######3   | 2101/3300 [04:17<34:18,  1.72s/it, lr: 1.0e-04 loss: 1.314e-01]
anha:  64%|######3   | 2102/3300 [04:17<34:26,  1.73s/it, lr: 1.0e-04 loss: 1.314e-01]
anha:  64%|######3   | 2102/3300 [04:17<34:26,  1.73s/it, lr: 1.0e-04 loss: 1.314e-01]
anha:  64%|######3   | 2102/3300 [04:19<34:26,  1.73s/it, lr: 1.0e-04 loss: 4.969e-02]
anha:  64%|######3   | 2102/3300 [04:19<34:26,  1.73s/it, lr: 1.0e-04 loss: 4.969e-02]
anha:  64%|######3   | 2103/3300 [04:19<35:36,  1.78s/it, lr: 1.0e-04 loss: 4.969e-02]
anha:  64%|######3   | 2103/3300 [04:19<35:36,  1.78s/it, lr: 1.0e-04 loss: 4.969e-02]
anha:  64%|######3   | 2103/3300 [04:21<35:36,  1.78s/it, lr: 1.0e-04 loss: 7.971e-02]
anha:  64%|######3   | 2103/3300 [04:21<35:36,  1.78s/it, lr: 1.0e-04 loss: 7.971e-02]
anha:  64%|######3   | 2104/3300 [04:21<36:52,  1.85s/it, lr: 1.0e-04 loss: 7.971e-02]
anha:  64%|######3   | 2104/3300 [04:21<36:52,  1.85s/it, lr: 1.0e-04 loss: 7.971e-02]
anha:  64%|######3   | 2104/3300 [04:25<36:52,  1.85s/it, lr: 1.0e-04 loss: 3.742e-02]
anha:  64%|######3   | 2104/3300 [04:25<36:52,  1.85s/it, lr: 1.0e-04 loss: 3.742e-02]
anha:  64%|######3   | 2105/3300 [04:25<45:17,  2.27s/it, lr: 1.0e-04 loss: 3.742e-02]
anha:  64%|######3   | 2105/3300 [04:25<45:17,  2.27s/it, lr: 1.0e-04 loss: 3.742e-02]
anha:  64%|######3   | 2105/3300 [04:27<45:17,  2.27s/it, lr: 1.0e-04 loss: 4.020e-02]
anha:  64%|######3   | 2105/3300 [04:27<45:17,  2.27s/it, lr: 1.0e-04 loss: 4.020e-02]
anha:  64%|######3   | 2106/3300 [04:27<43:34,  2.19s/it, lr: 1.0e-04 loss: 4.020e-02]
anha:  64%|######3   | 2106/3300 [04:27<43:34,  2.19s/it, lr: 1.0e-04 loss: 4.020e-02]
anha:  64%|######3   | 2106/3300 [04:28<43:34,  2.19s/it, lr: 1.0e-04 loss: 5.064e-02]
anha:  64%|######3   | 2106/3300 [04:28<43:34,  2.19s/it, lr: 1.0e-04 loss: 5.064e-02]
anha:  64%|######3   | 2107/3300 [04:28<36:40,  1.84s/it, lr: 1.0e-04 loss: 5.064e-02]
anha:  64%|######3   | 2107/3300 [04:28<36:40,  1.84s/it, lr: 1.0e-04 loss: 5.064e-02]
anha:  64%|######3   | 2107/3300 [04:31<36:40,  1.84s/it, lr: 1.0e-04 loss: 5.865e-02]
anha:  64%|######3   | 2107/3300 [04:31<36:40,  1.84s/it, lr: 1.0e-04 loss: 5.865e-02]
anha:  64%|######3   | 2108/3300 [04:31<45:55,  2.31s/it, lr: 1.0e-04 loss: 5.865e-02]
anha:  64%|######3   | 2108/3300 [04:31<45:55,  2.31s/it, lr: 1.0e-04 loss: 5.865e-02]
anha:  64%|######3   | 2108/3300 [04:34<45:55,  2.31s/it, lr: 1.0e-04 loss: 3.758e-02]
anha:  64%|######3   | 2108/3300 [04:34<45:55,  2.31s/it, lr: 1.0e-04 loss: 3.758e-02]
anha:  64%|######3   | 2109/3300 [04:34<47:02,  2.37s/it, lr: 1.0e-04 loss: 3.758e-02]
anha:  64%|######3   | 2109/3300 [04:34<47:02,  2.37s/it, lr: 1.0e-04 loss: 3.758e-02]
anha:  64%|######3   | 2109/3300 [04:35<47:02,  2.37s/it, lr: 1.0e-04 loss: 7.179e-02]
anha:  64%|######3   | 2109/3300 [04:35<47:02,  2.37s/it, lr: 1.0e-04 loss: 7.179e-02]
anha:  64%|######3   | 2110/3300 [04:36<46:59,  2.37s/it, lr: 1.0e-04 loss: 6.881e-02]
anha:  64%|######3   | 2110/3300 [04:36<46:59,  2.37s/it, lr: 1.0e-04 loss: 6.881e-02]
anha:  64%|######3   | 2111/3300 [04:36<29:58,  1.51s/it, lr: 1.0e-04 loss: 6.881e-02]
anha:  64%|######3   | 2111/3300 [04:36<29:58,  1.51s/it, lr: 1.0e-04 loss: 6.881e-02]
anha:  64%|######3   | 2111/3300 [04:38<29:58,  1.51s/it, lr: 1.0e-04 loss: 5.217e-02]
anha:  64%|######3   | 2111/3300 [04:38<29:58,  1.51s/it, lr: 1.0e-04 loss: 5.217e-02]
anha:  64%|######4   | 2112/3300 [04:38<32:12,  1.63s/it, lr: 1.0e-04 loss: 5.217e-02]
anha:  64%|######4   | 2112/3300 [04:38<32:12,  1.63s/it, lr: 1.0e-04 loss: 5.217e-02]
anha:  64%|######4   | 2112/3300 [04:39<32:12,  1.63s/it, lr: 1.0e-04 loss: 7.552e-02]
anha:  64%|######4   | 2112/3300 [04:39<32:12,  1.63s/it, lr: 1.0e-04 loss: 7.552e-02]
anha:  64%|######4   | 2113/3300 [04:39<29:44,  1.50s/it, lr: 1.0e-04 loss: 7.552e-02]
anha:  64%|######4   | 2113/3300 [04:39<29:44,  1.50s/it, lr: 1.0e-04 loss: 7.552e-02]
anha:  64%|######4   | 2113/3300 [04:41<29:44,  1.50s/it, lr: 1.0e-04 loss: 8.782e-02]
anha:  64%|######4   | 2113/3300 [04:41<29:44,  1.50s/it, lr: 1.0e-04 loss: 8.782e-02]
anha:  64%|######4   | 2114/3300 [04:41<32:15,  1.63s/it, lr: 1.0e-04 loss: 8.782e-02]
anha:  64%|######4   | 2114/3300 [04:41<32:15,  1.63s/it, lr: 1.0e-04 loss: 8.782e-02]
anha:  64%|######4   | 2114/3300 [04:43<32:15,  1.63s/it, lr: 1.0e-04 loss: 3.427e-02]
anha:  64%|######4   | 2114/3300 [04:43<32:15,  1.63s/it, lr: 1.0e-04 loss: 3.427e-02]
anha:  64%|######4   | 2115/3300 [04:43<33:59,  1.72s/it, lr: 1.0e-04 loss: 3.427e-02]
anha:  64%|######4   | 2115/3300 [04:43<33:59,  1.72s/it, lr: 1.0e-04 loss: 3.427e-02]
anha:  64%|######4   | 2115/3300 [04:44<33:59,  1.72s/it, lr: 1.0e-04 loss: 7.343e-02]
anha:  64%|######4   | 2115/3300 [04:44<33:59,  1.72s/it, lr: 1.0e-04 loss: 7.343e-02]
anha:  64%|######4   | 2116/3300 [04:44<30:00,  1.52s/it, lr: 1.0e-04 loss: 7.343e-02]
anha:  64%|######4   | 2116/3300 [04:44<30:00,  1.52s/it, lr: 1.0e-04 loss: 7.343e-02]
anha:  64%|######4   | 2116/3300 [04:46<30:00,  1.52s/it, lr: 1.0e-04 loss: 2.347e-01]
anha:  64%|######4   | 2116/3300 [04:46<30:00,  1.52s/it, lr: 1.0e-04 loss: 2.347e-01]
anha:  64%|######4   | 2117/3300 [04:46<31:05,  1.58s/it, lr: 1.0e-04 loss: 2.347e-01]
anha:  64%|######4   | 2117/3300 [04:46<31:05,  1.58s/it, lr: 1.0e-04 loss: 2.347e-01]
anha:  64%|######4   | 2117/3300 [04:48<31:05,  1.58s/it, lr: 1.0e-04 loss: 1.241e-01]
anha:  64%|######4   | 2117/3300 [04:48<31:05,  1.58s/it, lr: 1.0e-04 loss: 1.241e-01]
anha:  64%|######4   | 2118/3300 [04:48<36:25,  1.85s/it, lr: 1.0e-04 loss: 1.241e-01]
anha:  64%|######4   | 2118/3300 [04:48<36:25,  1.85s/it, lr: 1.0e-04 loss: 1.241e-01]
anha:  64%|######4   | 2118/3300 [04:51<36:25,  1.85s/it, lr: 1.0e-04 loss: 1.954e-01]
anha:  64%|######4   | 2118/3300 [04:51<36:25,  1.85s/it, lr: 1.0e-04 loss: 1.954e-01]
anha:  64%|######4   | 2119/3300 [04:51<44:43,  2.27s/it, lr: 1.0e-04 loss: 1.954e-01]
anha:  64%|######4   | 2119/3300 [04:51<44:43,  2.27s/it, lr: 1.0e-04 loss: 1.954e-01]
anha:  64%|######4   | 2119/3300 [04:53<44:43,  2.27s/it, lr: 1.0e-04 loss: 5.744e-02]
anha:  64%|######4   | 2119/3300 [04:53<44:43,  2.27s/it, lr: 1.0e-04 loss: 5.744e-02]
anha:  64%|######4   | 2120/3300 [04:54<44:41,  2.27s/it, lr: 1.0e-04 loss: 6.678e-02]
anha:  64%|######4   | 2120/3300 [04:54<44:41,  2.27s/it, lr: 1.0e-04 loss: 6.678e-02]
anha:  64%|######4   | 2121/3300 [04:54<28:41,  1.46s/it, lr: 1.0e-04 loss: 6.678e-02]
anha:  64%|######4   | 2121/3300 [04:54<28:41,  1.46s/it, lr: 1.0e-04 loss: 6.678e-02]
anha:  64%|######4   | 2121/3300 [04:56<28:41,  1.46s/it, lr: 1.0e-04 loss: 4.229e-02]
anha:  64%|######4   | 2121/3300 [04:56<28:41,  1.46s/it, lr: 1.0e-04 loss: 4.229e-02]
anha:  64%|######4   | 2122/3300 [04:56<30:21,  1.55s/it, lr: 1.0e-04 loss: 4.229e-02]
anha:  64%|######4   | 2122/3300 [04:56<30:21,  1.55s/it, lr: 1.0e-04 loss: 4.229e-02]
anha:  64%|######4   | 2122/3300 [04:58<30:21,  1.55s/it, lr: 1.0e-04 loss: 7.279e-02]
anha:  64%|######4   | 2122/3300 [04:58<30:21,  1.55s/it, lr: 1.0e-04 loss: 7.279e-02]
anha:  64%|######4   | 2123/3300 [04:58<32:28,  1.66s/it, lr: 1.0e-04 loss: 7.279e-02]
anha:  64%|######4   | 2123/3300 [04:58<32:28,  1.66s/it, lr: 1.0e-04 loss: 7.279e-02]
anha:  64%|######4   | 2123/3300 [04:59<32:28,  1.66s/it, lr: 1.0e-04 loss: 6.069e-02]
anha:  64%|######4   | 2123/3300 [04:59<32:28,  1.66s/it, lr: 1.0e-04 loss: 6.069e-02]
anha:  64%|######4   | 2124/3300 [04:59<28:59,  1.48s/it, lr: 1.0e-04 loss: 6.069e-02]
anha:  64%|######4   | 2124/3300 [04:59<28:59,  1.48s/it, lr: 1.0e-04 loss: 6.069e-02]
anha:  64%|######4   | 2124/3300 [05:01<28:59,  1.48s/it, lr: 1.0e-04 loss: 4.163e-02]
anha:  64%|######4   | 2124/3300 [05:01<28:59,  1.48s/it, lr: 1.0e-04 loss: 4.163e-02]
anha:  64%|######4   | 2125/3300 [05:01<31:33,  1.61s/it, lr: 1.0e-04 loss: 4.163e-02]
anha:  64%|######4   | 2125/3300 [05:01<31:33,  1.61s/it, lr: 1.0e-04 loss: 4.163e-02]
anha:  64%|######4   | 2125/3300 [05:02<31:33,  1.61s/it, lr: 1.0e-04 loss: 6.413e-02]
anha:  64%|######4   | 2125/3300 [05:02<31:33,  1.61s/it, lr: 1.0e-04 loss: 6.413e-02]
anha:  64%|######4   | 2126/3300 [05:02<28:10,  1.44s/it, lr: 1.0e-04 loss: 6.413e-02]
anha:  64%|######4   | 2126/3300 [05:02<28:10,  1.44s/it, lr: 1.0e-04 loss: 6.413e-02]
anha:  64%|######4   | 2126/3300 [05:04<28:10,  1.44s/it, lr: 1.0e-04 loss: 6.183e-02]
anha:  64%|######4   | 2126/3300 [05:04<28:10,  1.44s/it, lr: 1.0e-04 loss: 6.183e-02]
anha:  64%|######4   | 2127/3300 [05:04<31:10,  1.59s/it, lr: 1.0e-04 loss: 6.183e-02]
anha:  64%|######4   | 2127/3300 [05:04<31:10,  1.59s/it, lr: 1.0e-04 loss: 6.183e-02]
anha:  64%|######4   | 2127/3300 [05:05<31:10,  1.59s/it, lr: 1.0e-04 loss: 6.796e-02]
anha:  64%|######4   | 2127/3300 [05:05<31:10,  1.59s/it, lr: 1.0e-04 loss: 6.796e-02]
anha:  64%|######4   | 2128/3300 [05:05<27:54,  1.43s/it, lr: 1.0e-04 loss: 6.796e-02]
anha:  64%|######4   | 2128/3300 [05:05<27:54,  1.43s/it, lr: 1.0e-04 loss: 6.796e-02]
anha:  64%|######4   | 2128/3300 [05:07<27:54,  1.43s/it, lr: 1.0e-04 loss: 3.894e-02]
anha:  64%|######4   | 2128/3300 [05:07<27:54,  1.43s/it, lr: 1.0e-04 loss: 3.894e-02]
anha:  65%|######4   | 2129/3300 [05:07<29:30,  1.51s/it, lr: 1.0e-04 loss: 3.894e-02]
anha:  65%|######4   | 2129/3300 [05:07<29:30,  1.51s/it, lr: 1.0e-04 loss: 3.894e-02]
anha:  65%|######4   | 2129/3300 [05:10<29:30,  1.51s/it, lr: 1.0e-04 loss: 6.005e-02]
anha:  65%|######4   | 2129/3300 [05:10<29:30,  1.51s/it, lr: 1.0e-04 loss: 6.005e-02]
anha:  65%|######4   | 2130/3300 [05:14<29:29,  1.51s/it, lr: 1.0e-04 loss: 6.631e-02]
anha:  65%|######4   | 2130/3300 [05:14<29:29,  1.51s/it, lr: 1.0e-04 loss: 6.631e-02]
anha:  65%|######4   | 2131/3300 [05:14<31:31,  1.62s/it, lr: 1.0e-04 loss: 6.631e-02]
anha:  65%|######4   | 2131/3300 [05:14<31:31,  1.62s/it, lr: 1.0e-04 loss: 6.631e-02]
anha:  65%|######4   | 2131/3300 [05:16<31:31,  1.62s/it, lr: 1.0e-04 loss: 5.328e-02]
anha:  65%|######4   | 2131/3300 [05:16<31:31,  1.62s/it, lr: 1.0e-04 loss: 5.328e-02]
anha:  65%|######4   | 2132/3300 [05:16<33:35,  1.73s/it, lr: 1.0e-04 loss: 5.328e-02]
anha:  65%|######4   | 2132/3300 [05:16<33:35,  1.73s/it, lr: 1.0e-04 loss: 5.328e-02]
anha:  65%|######4   | 2132/3300 [05:17<33:35,  1.73s/it, lr: 1.0e-04 loss: 4.424e-02]
anha:  65%|######4   | 2132/3300 [05:17<33:35,  1.73s/it, lr: 1.0e-04 loss: 4.424e-02]
anha:  65%|######4   | 2133/3300 [05:17<29:57,  1.54s/it, lr: 1.0e-04 loss: 4.424e-02]
anha:  65%|######4   | 2133/3300 [05:17<29:57,  1.54s/it, lr: 1.0e-04 loss: 4.424e-02]
anha:  65%|######4   | 2133/3300 [05:19<29:57,  1.54s/it, lr: 1.0e-04 loss: 6.787e-02]
anha:  65%|######4   | 2133/3300 [05:19<29:57,  1.54s/it, lr: 1.0e-04 loss: 6.787e-02]
anha:  65%|######4   | 2134/3300 [05:19<32:06,  1.65s/it, lr: 1.0e-04 loss: 6.787e-02]
anha:  65%|######4   | 2134/3300 [05:19<32:06,  1.65s/it, lr: 1.0e-04 loss: 6.787e-02]
anha:  65%|######4   | 2134/3300 [05:22<32:06,  1.65s/it, lr: 1.0e-04 loss: 4.155e-02]
anha:  65%|######4   | 2134/3300 [05:22<32:06,  1.65s/it, lr: 1.0e-04 loss: 4.155e-02]
anha:  65%|######4   | 2135/3300 [05:22<41:47,  2.15s/it, lr: 1.0e-04 loss: 4.155e-02]
anha:  65%|######4   | 2135/3300 [05:22<41:47,  2.15s/it, lr: 1.0e-04 loss: 4.155e-02]
anha:  65%|######4   | 2135/3300 [05:24<41:47,  2.15s/it, lr: 1.0e-04 loss: 5.776e-02]
anha:  65%|######4   | 2135/3300 [05:24<41:47,  2.15s/it, lr: 1.0e-04 loss: 5.776e-02]
anha:  65%|######4   | 2136/3300 [05:24<40:38,  2.10s/it, lr: 1.0e-04 loss: 5.776e-02]
anha:  65%|######4   | 2136/3300 [05:24<40:38,  2.10s/it, lr: 1.0e-04 loss: 5.776e-02]
anha:  65%|######4   | 2136/3300 [05:26<40:38,  2.10s/it, lr: 1.0e-04 loss: 4.393e-02]
anha:  65%|######4   | 2136/3300 [05:26<40:38,  2.10s/it, lr: 1.0e-04 loss: 4.393e-02]
anha:  65%|######4   | 2137/3300 [05:26<38:37,  1.99s/it, lr: 1.0e-04 loss: 4.393e-02]
anha:  65%|######4   | 2137/3300 [05:26<38:37,  1.99s/it, lr: 1.0e-04 loss: 4.393e-02]
anha:  65%|######4   | 2137/3300 [05:27<38:37,  1.99s/it, lr: 1.0e-04 loss: 5.593e-02]
anha:  65%|######4   | 2137/3300 [05:27<38:37,  1.99s/it, lr: 1.0e-04 loss: 5.593e-02]
anha:  65%|######4   | 2138/3300 [05:27<32:58,  1.70s/it, lr: 1.0e-04 loss: 5.593e-02]
anha:  65%|######4   | 2138/3300 [05:27<32:58,  1.70s/it, lr: 1.0e-04 loss: 5.593e-02]
anha:  65%|######4   | 2138/3300 [05:30<32:58,  1.70s/it, lr: 1.0e-04 loss: 2.290e-01]
anha:  65%|######4   | 2138/3300 [05:30<32:58,  1.70s/it, lr: 1.0e-04 loss: 2.290e-01]
anha:  65%|######4   | 2139/3300 [05:30<41:58,  2.17s/it, lr: 1.0e-04 loss: 2.290e-01]
anha:  65%|######4   | 2139/3300 [05:30<41:58,  2.17s/it, lr: 1.0e-04 loss: 2.290e-01]
anha:  65%|######4   | 2139/3300 [05:34<41:58,  2.17s/it, lr: 1.0e-04 loss: 4.249e-02]
anha:  65%|######4   | 2139/3300 [05:34<41:58,  2.17s/it, lr: 1.0e-04 loss: 4.249e-02]
anha:  65%|######4   | 2140/3300 [05:35<41:56,  2.17s/it, lr: 1.0e-04 loss: 2.241e-01]
anha:  65%|######4   | 2140/3300 [05:35<41:56,  2.17s/it, lr: 1.0e-04 loss: 2.241e-01]
anha:  65%|######4   | 2141/3300 [05:35<27:38,  1.43s/it, lr: 1.0e-04 loss: 2.241e-01]
anha:  65%|######4   | 2141/3300 [05:35<27:38,  1.43s/it, lr: 1.0e-04 loss: 2.241e-01]
anha:  65%|######4   | 2141/3300 [05:37<27:38,  1.43s/it, lr: 1.0e-04 loss: 2.282e-01]
anha:  65%|######4   | 2141/3300 [05:37<27:38,  1.43s/it, lr: 1.0e-04 loss: 2.282e-01]
anha:  65%|######4   | 2142/3300 [05:37<30:13,  1.57s/it, lr: 1.0e-04 loss: 2.282e-01]
anha:  65%|######4   | 2142/3300 [05:37<30:13,  1.57s/it, lr: 1.0e-04 loss: 2.282e-01]
anha:  65%|######4   | 2142/3300 [05:40<30:13,  1.57s/it, lr: 1.0e-04 loss: 2.231e-01]
anha:  65%|######4   | 2142/3300 [05:40<30:13,  1.57s/it, lr: 1.0e-04 loss: 2.231e-01]
anha:  65%|######4   | 2143/3300 [05:40<39:54,  2.07s/it, lr: 1.0e-04 loss: 2.231e-01]
anha:  65%|######4   | 2143/3300 [05:40<39:54,  2.07s/it, lr: 1.0e-04 loss: 2.231e-01]
anha:  65%|######4   | 2143/3300 [05:41<39:54,  2.07s/it, lr: 1.0e-04 loss: 1.614e-01]
anha:  65%|######4   | 2143/3300 [05:41<39:54,  2.07s/it, lr: 1.0e-04 loss: 1.614e-01]
anha:  65%|######4   | 2144/3300 [05:41<34:19,  1.78s/it, lr: 1.0e-04 loss: 1.614e-01]
anha:  65%|######4   | 2144/3300 [05:41<34:19,  1.78s/it, lr: 1.0e-04 loss: 1.614e-01]
anha:  65%|######4   | 2144/3300 [05:44<34:19,  1.78s/it, lr: 1.0e-04 loss: 3.433e-02]
anha:  65%|######4   | 2144/3300 [05:44<34:19,  1.78s/it, lr: 1.0e-04 loss: 3.433e-02]
anha:  65%|######5   | 2145/3300 [05:44<42:49,  2.22s/it, lr: 1.0e-04 loss: 3.433e-02]
anha:  65%|######5   | 2145/3300 [05:44<42:49,  2.22s/it, lr: 1.0e-04 loss: 3.433e-02]
anha:  65%|######5   | 2145/3300 [05:45<42:49,  2.22s/it, lr: 1.0e-04 loss: 5.045e-02]
anha:  65%|######5   | 2145/3300 [05:45<42:49,  2.22s/it, lr: 1.0e-04 loss: 5.045e-02]
anha:  65%|######5   | 2146/3300 [05:45<36:06,  1.88s/it, lr: 1.0e-04 loss: 5.045e-02]
anha:  65%|######5   | 2146/3300 [05:45<36:06,  1.88s/it, lr: 1.0e-04 loss: 5.045e-02]
anha:  65%|######5   | 2146/3300 [05:46<36:06,  1.88s/it, lr: 1.0e-04 loss: 9.428e-02]
anha:  65%|######5   | 2146/3300 [05:46<36:06,  1.88s/it, lr: 1.0e-04 loss: 9.428e-02]
anha:  65%|######5   | 2147/3300 [05:46<31:14,  1.63s/it, lr: 1.0e-04 loss: 9.428e-02]
anha:  65%|######5   | 2147/3300 [05:46<31:14,  1.63s/it, lr: 1.0e-04 loss: 9.428e-02]
anha:  65%|######5   | 2147/3300 [05:48<31:14,  1.63s/it, lr: 1.0e-04 loss: 3.885e-02]
anha:  65%|######5   | 2147/3300 [05:48<31:14,  1.63s/it, lr: 1.0e-04 loss: 3.885e-02]
anha:  65%|######5   | 2148/3300 [05:48<33:03,  1.72s/it, lr: 1.0e-04 loss: 3.885e-02]
anha:  65%|######5   | 2148/3300 [05:48<33:03,  1.72s/it, lr: 1.0e-04 loss: 3.885e-02]
anha:  65%|######5   | 2148/3300 [05:49<33:03,  1.72s/it, lr: 1.0e-04 loss: 7.875e-02]
anha:  65%|######5   | 2148/3300 [05:49<33:03,  1.72s/it, lr: 1.0e-04 loss: 7.875e-02]
anha:  65%|######5   | 2149/3300 [05:49<28:54,  1.51s/it, lr: 1.0e-04 loss: 7.875e-02]
anha:  65%|######5   | 2149/3300 [05:49<28:54,  1.51s/it, lr: 1.0e-04 loss: 7.875e-02]
anha:  65%|######5   | 2149/3300 [05:51<28:54,  1.51s/it, lr: 1.0e-04 loss: 7.602e-02]
anha:  65%|######5   | 2149/3300 [05:51<28:54,  1.51s/it, lr: 1.0e-04 loss: 7.602e-02]
anha:  65%|######5   | 2150/3300 [05:54<28:52,  1.51s/it, lr: 1.0e-04 loss: 5.305e-02]
anha:  65%|######5   | 2150/3300 [05:54<28:52,  1.51s/it, lr: 1.0e-04 loss: 5.305e-02]
anha:  65%|######5   | 2151/3300 [05:54<30:53,  1.61s/it, lr: 1.0e-04 loss: 5.305e-02]
anha:  65%|######5   | 2151/3300 [05:54<30:53,  1.61s/it, lr: 1.0e-04 loss: 5.305e-02]
anha:  65%|######5   | 2151/3300 [05:55<30:53,  1.61s/it, lr: 1.0e-04 loss: 4.464e-02]
anha:  65%|######5   | 2151/3300 [05:55<30:53,  1.61s/it, lr: 1.0e-04 loss: 4.464e-02]
anha:  65%|######5   | 2152/3300 [05:55<28:01,  1.46s/it, lr: 1.0e-04 loss: 4.464e-02]
anha:  65%|######5   | 2152/3300 [05:55<28:01,  1.46s/it, lr: 1.0e-04 loss: 4.464e-02]
anha:  65%|######5   | 2152/3300 [05:57<28:01,  1.46s/it, lr: 1.0e-04 loss: 1.199e-01]
anha:  65%|######5   | 2152/3300 [05:57<28:01,  1.46s/it, lr: 1.0e-04 loss: 1.199e-01]
anha:  65%|######5   | 2153/3300 [05:57<29:21,  1.54s/it, lr: 1.0e-04 loss: 1.199e-01]
anha:  65%|######5   | 2153/3300 [05:57<29:21,  1.54s/it, lr: 1.0e-04 loss: 1.199e-01]
anha:  65%|######5   | 2153/3300 [05:58<29:21,  1.54s/it, lr: 1.0e-04 loss: 1.641e-01]
anha:  65%|######5   | 2153/3300 [05:58<29:21,  1.54s/it, lr: 1.0e-04 loss: 1.641e-01]
anha:  65%|######5   | 2154/3300 [05:58<26:32,  1.39s/it, lr: 1.0e-04 loss: 1.641e-01]
anha:  65%|######5   | 2154/3300 [05:58<26:32,  1.39s/it, lr: 1.0e-04 loss: 1.641e-01]
anha:  65%|######5   | 2154/3300 [06:01<26:32,  1.39s/it, lr: 1.0e-04 loss: 1.097e-01]
anha:  65%|######5   | 2154/3300 [06:01<26:32,  1.39s/it, lr: 1.0e-04 loss: 1.097e-01]
anha:  65%|######5   | 2155/3300 [06:01<37:38,  1.97s/it, lr: 1.0e-04 loss: 1.097e-01]
anha:  65%|######5   | 2155/3300 [06:01<37:38,  1.97s/it, lr: 1.0e-04 loss: 1.097e-01]
anha:  65%|######5   | 2155/3300 [06:05<37:38,  1.97s/it, lr: 1.0e-04 loss: 4.958e-02]
anha:  65%|######5   | 2155/3300 [06:05<37:38,  1.97s/it, lr: 1.0e-04 loss: 4.958e-02]
anha:  65%|######5   | 2156/3300 [06:05<44:36,  2.34s/it, lr: 1.0e-04 loss: 4.958e-02]
anha:  65%|######5   | 2156/3300 [06:05<44:36,  2.34s/it, lr: 1.0e-04 loss: 4.958e-02]
anha:  65%|######5   | 2156/3300 [06:06<44:36,  2.34s/it, lr: 1.0e-04 loss: 3.521e-02]
anha:  65%|######5   | 2156/3300 [06:06<44:36,  2.34s/it, lr: 1.0e-04 loss: 3.521e-02]
anha:  65%|######5   | 2157/3300 [06:06<42:26,  2.23s/it, lr: 1.0e-04 loss: 3.521e-02]
anha:  65%|######5   | 2157/3300 [06:06<42:26,  2.23s/it, lr: 1.0e-04 loss: 3.521e-02]
anha:  65%|######5   | 2157/3300 [06:07<42:26,  2.23s/it, lr: 1.0e-04 loss: 4.773e-02]
anha:  65%|######5   | 2157/3300 [06:07<42:26,  2.23s/it, lr: 1.0e-04 loss: 4.773e-02]
anha:  65%|######5   | 2158/3300 [06:07<35:37,  1.87s/it, lr: 1.0e-04 loss: 4.773e-02]
anha:  65%|######5   | 2158/3300 [06:07<35:37,  1.87s/it, lr: 1.0e-04 loss: 4.773e-02]
anha:  65%|######5   | 2158/3300 [06:11<35:37,  1.87s/it, lr: 1.0e-04 loss: 3.910e-02]
anha:  65%|######5   | 2158/3300 [06:11<35:37,  1.87s/it, lr: 1.0e-04 loss: 3.910e-02]
anha:  65%|######5   | 2159/3300 [06:11<43:59,  2.31s/it, lr: 1.0e-04 loss: 3.910e-02]
anha:  65%|######5   | 2159/3300 [06:11<43:59,  2.31s/it, lr: 1.0e-04 loss: 3.910e-02]
anha:  65%|######5   | 2159/3300 [06:14<43:59,  2.31s/it, lr: 1.0e-04 loss: 6.254e-02]
anha:  65%|######5   | 2159/3300 [06:14<43:59,  2.31s/it, lr: 1.0e-04 loss: 6.254e-02]
anha:  65%|######5   | 2160/3300 [06:17<43:57,  2.31s/it, lr: 1.0e-04 loss: 1.113e-01]
anha:  65%|######5   | 2160/3300 [06:17<43:57,  2.31s/it, lr: 1.0e-04 loss: 1.113e-01]
anha:  65%|######5   | 2161/3300 [06:17<37:52,  2.00s/it, lr: 1.0e-04 loss: 1.113e-01]
anha:  65%|######5   | 2161/3300 [06:17<37:52,  2.00s/it, lr: 1.0e-04 loss: 1.113e-01]
anha:  65%|######5   | 2161/3300 [06:18<37:52,  2.00s/it, lr: 1.0e-04 loss: 3.850e-02]
anha:  65%|######5   | 2161/3300 [06:18<37:52,  2.00s/it, lr: 1.0e-04 loss: 3.850e-02]
anha:  66%|######5   | 2162/3300 [06:18<33:12,  1.75s/it, lr: 1.0e-04 loss: 3.850e-02]
anha:  66%|######5   | 2162/3300 [06:18<33:12,  1.75s/it, lr: 1.0e-04 loss: 3.850e-02]
anha:  66%|######5   | 2162/3300 [06:20<33:12,  1.75s/it, lr: 1.0e-04 loss: 1.554e-01]
anha:  66%|######5   | 2162/3300 [06:20<33:12,  1.75s/it, lr: 1.0e-04 loss: 1.554e-01]
anha:  66%|######5   | 2163/3300 [06:20<34:10,  1.80s/it, lr: 1.0e-04 loss: 1.554e-01]
anha:  66%|######5   | 2163/3300 [06:20<34:10,  1.80s/it, lr: 1.0e-04 loss: 1.554e-01]
anha:  66%|######5   | 2163/3300 [06:24<34:10,  1.80s/it, lr: 1.0e-04 loss: 1.101e-01]
anha:  66%|######5   | 2163/3300 [06:24<34:10,  1.80s/it, lr: 1.0e-04 loss: 1.101e-01]
anha:  66%|######5   | 2164/3300 [06:24<42:11,  2.23s/it, lr: 1.0e-04 loss: 1.101e-01]
anha:  66%|######5   | 2164/3300 [06:24<42:11,  2.23s/it, lr: 1.0e-04 loss: 1.101e-01]
anha:  66%|######5   | 2164/3300 [06:25<42:11,  2.23s/it, lr: 1.0e-04 loss: 4.136e-02]
anha:  66%|######5   | 2164/3300 [06:25<42:11,  2.23s/it, lr: 1.0e-04 loss: 4.136e-02]
anha:  66%|######5   | 2165/3300 [06:25<35:37,  1.88s/it, lr: 1.0e-04 loss: 4.136e-02]
anha:  66%|######5   | 2165/3300 [06:25<35:37,  1.88s/it, lr: 1.0e-04 loss: 4.136e-02]
anha:  66%|######5   | 2165/3300 [06:28<35:37,  1.88s/it, lr: 1.0e-04 loss: 5.929e-02]
anha:  66%|######5   | 2165/3300 [06:28<35:37,  1.88s/it, lr: 1.0e-04 loss: 5.929e-02]
anha:  66%|######5   | 2166/3300 [06:28<43:59,  2.33s/it, lr: 1.0e-04 loss: 5.929e-02]
anha:  66%|######5   | 2166/3300 [06:28<43:59,  2.33s/it, lr: 1.0e-04 loss: 5.929e-02]
anha:  66%|######5   | 2166/3300 [06:30<43:59,  2.33s/it, lr: 1.0e-04 loss: 5.614e-02]
anha:  66%|######5   | 2166/3300 [06:30<43:59,  2.33s/it, lr: 1.0e-04 loss: 5.614e-02]
anha:  66%|######5   | 2167/3300 [06:30<40:40,  2.15s/it, lr: 1.0e-04 loss: 5.614e-02]
anha:  66%|######5   | 2167/3300 [06:30<40:40,  2.15s/it, lr: 1.0e-04 loss: 5.614e-02]
anha:  66%|######5   | 2167/3300 [06:33<40:40,  2.15s/it, lr: 1.0e-04 loss: 3.646e-02]
anha:  66%|######5   | 2167/3300 [06:33<40:40,  2.15s/it, lr: 1.0e-04 loss: 3.646e-02]
anha:  66%|######5   | 2168/3300 [06:33<46:10,  2.45s/it, lr: 1.0e-04 loss: 3.646e-02]
anha:  66%|######5   | 2168/3300 [06:33<46:10,  2.45s/it, lr: 1.0e-04 loss: 3.646e-02]
anha:  66%|######5   | 2168/3300 [06:35<46:10,  2.45s/it, lr: 1.0e-04 loss: 1.244e-01]
anha:  66%|######5   | 2168/3300 [06:35<46:10,  2.45s/it, lr: 1.0e-04 loss: 1.244e-01]
anha:  66%|######5   | 2169/3300 [06:35<43:52,  2.33s/it, lr: 1.0e-04 loss: 1.244e-01]
anha:  66%|######5   | 2169/3300 [06:35<43:52,  2.33s/it, lr: 1.0e-04 loss: 1.244e-01]
anha:  66%|######5   | 2169/3300 [06:36<43:52,  2.33s/it, lr: 1.0e-04 loss: 5.545e-02]
anha:  66%|######5   | 2169/3300 [06:36<43:52,  2.33s/it, lr: 1.0e-04 loss: 5.545e-02]
anha:  66%|######5   | 2170/3300 [06:38<43:50,  2.33s/it, lr: 1.0e-04 loss: 5.176e-02]
anha:  66%|######5   | 2170/3300 [06:38<43:50,  2.33s/it, lr: 1.0e-04 loss: 5.176e-02]
anha:  66%|######5   | 2171/3300 [06:38<31:12,  1.66s/it, lr: 1.0e-04 loss: 5.176e-02]
anha:  66%|######5   | 2171/3300 [06:38<31:12,  1.66s/it, lr: 1.0e-04 loss: 5.176e-02]
anha:  66%|######5   | 2171/3300 [06:41<31:12,  1.66s/it, lr: 1.0e-04 loss: 4.380e-02]
anha:  66%|######5   | 2171/3300 [06:41<31:12,  1.66s/it, lr: 1.0e-04 loss: 4.380e-02]
anha:  66%|######5   | 2172/3300 [06:41<39:05,  2.08s/it, lr: 1.0e-04 loss: 4.380e-02]
anha:  66%|######5   | 2172/3300 [06:41<39:05,  2.08s/it, lr: 1.0e-04 loss: 4.380e-02]
anha:  66%|######5   | 2172/3300 [06:45<39:05,  2.08s/it, lr: 1.0e-04 loss: 1.300e-01]
anha:  66%|######5   | 2172/3300 [06:45<39:05,  2.08s/it, lr: 1.0e-04 loss: 1.300e-01]
anha:  66%|######5   | 2173/3300 [06:45<45:58,  2.45s/it, lr: 1.0e-04 loss: 1.300e-01]
anha:  66%|######5   | 2173/3300 [06:45<45:58,  2.45s/it, lr: 1.0e-04 loss: 1.300e-01]
anha:  66%|######5   | 2173/3300 [06:48<45:58,  2.45s/it, lr: 1.0e-04 loss: 6.963e-02]
anha:  66%|######5   | 2173/3300 [06:48<45:58,  2.45s/it, lr: 1.0e-04 loss: 6.963e-02]
anha:  66%|######5   | 2174/3300 [06:48<50:12,  2.68s/it, lr: 1.0e-04 loss: 6.963e-02]
anha:  66%|######5   | 2174/3300 [06:48<50:12,  2.68s/it, lr: 1.0e-04 loss: 6.963e-02]
anha:  66%|######5   | 2174/3300 [06:49<50:12,  2.68s/it, lr: 1.0e-04 loss: 7.866e-02]
anha:  66%|######5   | 2174/3300 [06:49<50:12,  2.68s/it, lr: 1.0e-04 loss: 7.866e-02]
anha:  66%|######5   | 2175/3300 [06:49<41:25,  2.21s/it, lr: 1.0e-04 loss: 7.866e-02]
anha:  66%|######5   | 2175/3300 [06:49<41:25,  2.21s/it, lr: 1.0e-04 loss: 7.866e-02]
anha:  66%|######5   | 2175/3300 [06:50<41:25,  2.21s/it, lr: 1.0e-04 loss: 8.914e-02]
anha:  66%|######5   | 2175/3300 [06:50<41:25,  2.21s/it, lr: 1.0e-04 loss: 8.914e-02]
anha:  66%|######5   | 2176/3300 [06:50<34:54,  1.86s/it, lr: 1.0e-04 loss: 8.914e-02]
anha:  66%|######5   | 2176/3300 [06:50<34:54,  1.86s/it, lr: 1.0e-04 loss: 8.914e-02]
anha:  66%|######5   | 2176/3300 [06:54<34:54,  1.86s/it, lr: 1.0e-04 loss: 9.911e-02]
anha:  66%|######5   | 2176/3300 [06:54<34:54,  1.86s/it, lr: 1.0e-04 loss: 9.911e-02]
anha:  66%|######5   | 2177/3300 [06:54<43:38,  2.33s/it, lr: 1.0e-04 loss: 9.911e-02]
anha:  66%|######5   | 2177/3300 [06:54<43:38,  2.33s/it, lr: 1.0e-04 loss: 9.911e-02]
anha:  66%|######5   | 2177/3300 [06:55<43:38,  2.33s/it, lr: 1.0e-04 loss: 5.859e-02]
anha:  66%|######5   | 2177/3300 [06:55<43:38,  2.33s/it, lr: 1.0e-04 loss: 5.859e-02]
anha:  66%|######6   | 2178/3300 [06:55<37:10,  1.99s/it, lr: 1.0e-04 loss: 5.859e-02]
anha:  66%|######6   | 2178/3300 [06:55<37:10,  1.99s/it, lr: 1.0e-04 loss: 5.859e-02]
anha:  66%|######6   | 2178/3300 [06:56<37:10,  1.99s/it, lr: 1.0e-04 loss: 5.518e-02]
anha:  66%|######6   | 2178/3300 [06:56<37:10,  1.99s/it, lr: 1.0e-04 loss: 5.518e-02]
anha:  66%|######6   | 2179/3300 [06:56<31:43,  1.70s/it, lr: 1.0e-04 loss: 5.518e-02]
anha:  66%|######6   | 2179/3300 [06:56<31:43,  1.70s/it, lr: 1.0e-04 loss: 5.518e-02]
anha:  66%|######6   | 2179/3300 [06:57<31:43,  1.70s/it, lr: 1.0e-04 loss: 5.976e-02]
anha:  66%|######6   | 2179/3300 [06:57<31:43,  1.70s/it, lr: 1.0e-04 loss: 5.976e-02]
anha:  66%|######6   | 2180/3300 [06:58<31:41,  1.70s/it, lr: 1.0e-04 loss: 4.814e-02]
anha:  66%|######6   | 2180/3300 [06:58<31:41,  1.70s/it, lr: 1.0e-04 loss: 4.814e-02]
anha:  66%|######6   | 2181/3300 [06:58<21:29,  1.15s/it, lr: 1.0e-04 loss: 4.814e-02]
anha:  66%|######6   | 2181/3300 [06:58<21:29,  1.15s/it, lr: 1.0e-04 loss: 4.814e-02]
anha:  66%|######6   | 2181/3300 [07:02<21:29,  1.15s/it, lr: 1.0e-04 loss: 4.823e-02]
anha:  66%|######6   | 2181/3300 [07:02<21:29,  1.15s/it, lr: 1.0e-04 loss: 4.823e-02]
anha:  66%|######6   | 2182/3300 [07:02<31:39,  1.70s/it, lr: 1.0e-04 loss: 4.823e-02]
anha:  66%|######6   | 2182/3300 [07:02<31:39,  1.70s/it, lr: 1.0e-04 loss: 4.823e-02]
anha:  66%|######6   | 2182/3300 [07:03<31:39,  1.70s/it, lr: 1.0e-04 loss: 1.767e-01]
anha:  66%|######6   | 2182/3300 [07:03<31:39,  1.70s/it, lr: 1.0e-04 loss: 1.767e-01]
anha:  66%|######6   | 2183/3300 [07:03<28:16,  1.52s/it, lr: 1.0e-04 loss: 1.767e-01]
anha:  66%|######6   | 2183/3300 [07:03<28:16,  1.52s/it, lr: 1.0e-04 loss: 1.767e-01]
anha:  66%|######6   | 2183/3300 [07:05<28:16,  1.52s/it, lr: 1.0e-04 loss: 4.939e-02]
anha:  66%|######6   | 2183/3300 [07:05<28:16,  1.52s/it, lr: 1.0e-04 loss: 4.939e-02]
anha:  66%|######6   | 2184/3300 [07:05<29:22,  1.58s/it, lr: 1.0e-04 loss: 4.939e-02]
anha:  66%|######6   | 2184/3300 [07:05<29:22,  1.58s/it, lr: 1.0e-04 loss: 4.939e-02]
anha:  66%|######6   | 2184/3300 [07:07<29:22,  1.58s/it, lr: 1.0e-04 loss: 6.021e-02]
anha:  66%|######6   | 2184/3300 [07:07<29:22,  1.58s/it, lr: 1.0e-04 loss: 6.021e-02]
anha:  66%|######6   | 2185/3300 [07:07<31:24,  1.69s/it, lr: 1.0e-04 loss: 6.021e-02]
anha:  66%|######6   | 2185/3300 [07:07<31:24,  1.69s/it, lr: 1.0e-04 loss: 6.021e-02]
anha:  66%|######6   | 2185/3300 [07:09<31:24,  1.69s/it, lr: 1.0e-04 loss: 4.576e-02]
anha:  66%|######6   | 2185/3300 [07:09<31:24,  1.69s/it, lr: 1.0e-04 loss: 4.576e-02]
anha:  66%|######6   | 2186/3300 [07:09<32:46,  1.77s/it, lr: 1.0e-04 loss: 4.576e-02]
anha:  66%|######6   | 2186/3300 [07:09<32:46,  1.77s/it, lr: 1.0e-04 loss: 4.576e-02]
anha:  66%|######6   | 2186/3300 [07:10<32:46,  1.77s/it, lr: 1.0e-04 loss: 2.237e-01]
anha:  66%|######6   | 2186/3300 [07:10<32:46,  1.77s/it, lr: 1.0e-04 loss: 2.237e-01]
anha:  66%|######6   | 2187/3300 [07:10<29:25,  1.59s/it, lr: 1.0e-04 loss: 2.237e-01]
anha:  66%|######6   | 2187/3300 [07:10<29:25,  1.59s/it, lr: 1.0e-04 loss: 2.237e-01]
anha:  66%|######6   | 2187/3300 [07:13<29:25,  1.59s/it, lr: 1.0e-04 loss: 9.630e-02]
anha:  66%|######6   | 2187/3300 [07:13<29:25,  1.59s/it, lr: 1.0e-04 loss: 9.630e-02]
anha:  66%|######6   | 2188/3300 [07:13<39:27,  2.13s/it, lr: 1.0e-04 loss: 9.630e-02]
anha:  66%|######6   | 2188/3300 [07:13<39:27,  2.13s/it, lr: 1.0e-04 loss: 9.630e-02]
anha:  66%|######6   | 2188/3300 [07:17<39:27,  2.13s/it, lr: 1.0e-04 loss: 1.263e-01]
anha:  66%|######6   | 2188/3300 [07:17<39:27,  2.13s/it, lr: 1.0e-04 loss: 1.263e-01]
anha:  66%|######6   | 2189/3300 [07:17<46:49,  2.53s/it, lr: 1.0e-04 loss: 1.263e-01]
anha:  66%|######6   | 2189/3300 [07:17<46:49,  2.53s/it, lr: 1.0e-04 loss: 1.263e-01]
anha:  66%|######6   | 2189/3300 [07:20<46:49,  2.53s/it, lr: 1.0e-04 loss: 5.282e-02]
anha:  66%|######6   | 2189/3300 [07:20<46:49,  2.53s/it, lr: 1.0e-04 loss: 5.282e-02]
anha:  66%|######6   | 2190/3300 [07:23<46:46,  2.53s/it, lr: 1.0e-04 loss: 7.408e-02]
anha:  66%|######6   | 2190/3300 [07:23<46:46,  2.53s/it, lr: 1.0e-04 loss: 7.408e-02]
anha:  66%|######6   | 2191/3300 [07:23<38:42,  2.09s/it, lr: 1.0e-04 loss: 7.408e-02]
anha:  66%|######6   | 2191/3300 [07:23<38:42,  2.09s/it, lr: 1.0e-04 loss: 7.408e-02]
anha:  66%|######6   | 2191/3300 [07:26<38:42,  2.09s/it, lr: 1.0e-04 loss: 1.534e-01]
anha:  66%|######6   | 2191/3300 [07:26<38:42,  2.09s/it, lr: 1.0e-04 loss: 1.534e-01]
anha:  66%|######6   | 2192/3300 [07:26<44:28,  2.41s/it, lr: 1.0e-04 loss: 1.534e-01]
anha:  66%|######6   | 2192/3300 [07:26<44:28,  2.41s/it, lr: 1.0e-04 loss: 1.534e-01]
anha:  66%|######6   | 2192/3300 [07:28<44:28,  2.41s/it, lr: 1.0e-04 loss: 2.705e-02]
anha:  66%|######6   | 2192/3300 [07:28<44:28,  2.41s/it, lr: 1.0e-04 loss: 2.705e-02]
anha:  66%|######6   | 2193/3300 [07:28<42:15,  2.29s/it, lr: 1.0e-04 loss: 2.705e-02]
anha:  66%|######6   | 2193/3300 [07:28<42:15,  2.29s/it, lr: 1.0e-04 loss: 2.705e-02]
anha:  66%|######6   | 2193/3300 [07:31<42:15,  2.29s/it, lr: 1.0e-04 loss: 3.710e-02]
anha:  66%|######6   | 2193/3300 [07:31<42:15,  2.29s/it, lr: 1.0e-04 loss: 3.710e-02]
anha:  66%|######6   | 2194/3300 [07:31<43:18,  2.35s/it, lr: 1.0e-04 loss: 3.710e-02]
anha:  66%|######6   | 2194/3300 [07:31<43:18,  2.35s/it, lr: 1.0e-04 loss: 3.710e-02]
anha:  66%|######6   | 2194/3300 [07:34<43:18,  2.35s/it, lr: 1.0e-04 loss: 8.095e-02]
anha:  66%|######6   | 2194/3300 [07:34<43:18,  2.35s/it, lr: 1.0e-04 loss: 8.095e-02]
anha:  67%|######6   | 2195/3300 [07:34<48:53,  2.65s/it, lr: 1.0e-04 loss: 8.095e-02]
anha:  67%|######6   | 2195/3300 [07:34<48:53,  2.65s/it, lr: 1.0e-04 loss: 8.095e-02]
anha:  67%|######6   | 2195/3300 [07:36<48:53,  2.65s/it, lr: 1.0e-04 loss: 6.055e-02]
anha:  67%|######6   | 2195/3300 [07:36<48:53,  2.65s/it, lr: 1.0e-04 loss: 6.055e-02]
anha:  67%|######6   | 2196/3300 [07:36<45:13,  2.46s/it, lr: 1.0e-04 loss: 6.055e-02]
anha:  67%|######6   | 2196/3300 [07:36<45:13,  2.46s/it, lr: 1.0e-04 loss: 6.055e-02]
anha:  67%|######6   | 2196/3300 [07:38<45:13,  2.46s/it, lr: 1.0e-04 loss: 5.399e-02]
anha:  67%|######6   | 2196/3300 [07:38<45:13,  2.46s/it, lr: 1.0e-04 loss: 5.399e-02]
anha:  67%|######6   | 2197/3300 [07:38<42:53,  2.33s/it, lr: 1.0e-04 loss: 5.399e-02]
anha:  67%|######6   | 2197/3300 [07:38<42:53,  2.33s/it, lr: 1.0e-04 loss: 5.399e-02]
anha:  67%|######6   | 2197/3300 [07:42<42:53,  2.33s/it, lr: 1.0e-04 loss: 9.280e-02]
anha:  67%|######6   | 2197/3300 [07:42<42:53,  2.33s/it, lr: 1.0e-04 loss: 9.280e-02]
anha:  67%|######6   | 2198/3300 [07:42<48:31,  2.64s/it, lr: 1.0e-04 loss: 9.280e-02]
anha:  67%|######6   | 2198/3300 [07:42<48:31,  2.64s/it, lr: 1.0e-04 loss: 9.280e-02]
anha:  67%|######6   | 2198/3300 [07:44<48:31,  2.64s/it, lr: 1.0e-04 loss: 5.612e-02]
anha:  67%|######6   | 2198/3300 [07:44<48:31,  2.64s/it, lr: 1.0e-04 loss: 5.612e-02]
anha:  67%|######6   | 2199/3300 [07:44<44:53,  2.45s/it, lr: 1.0e-04 loss: 5.612e-02]
anha:  67%|######6   | 2199/3300 [07:44<44:53,  2.45s/it, lr: 1.0e-04 loss: 5.612e-02]
anha:  67%|######6   | 2199/3300 [07:47<44:53,  2.45s/it, lr: 1.0e-04 loss: 4.403e-02]
anha:  67%|######6   | 2199/3300 [07:47<44:53,  2.45s/it, lr: 1.0e-04 loss: 4.403e-02]
anha:  67%|######6   | 2200/3300 [07:49<44:51,  2.45s/it, lr: 1.0e-04 loss: 5.592e-02]
anha:  67%|######6   | 2200/3300 [07:49<44:51,  2.45s/it, lr: 1.0e-04 loss: 5.592e-02]
anha:  67%|######6   | 2201/3300 [07:49<32:38,  1.78s/it, lr: 1.0e-04 loss: 5.592e-02]
anha:  67%|######6   | 2201/3300 [07:49<32:38,  1.78s/it, lr: 1.0e-04 loss: 5.592e-02]
anha:  67%|######6   | 2201/3300 [07:51<32:38,  1.78s/it, lr: 1.0e-04 loss: 1.020e-01]
anha:  67%|######6   | 2201/3300 [07:51<32:38,  1.78s/it, lr: 1.0e-04 loss: 1.020e-01]
anha:  67%|######6   | 2202/3300 [07:51<33:29,  1.83s/it, lr: 1.0e-04 loss: 1.020e-01]
anha:  67%|######6   | 2202/3300 [07:51<33:29,  1.83s/it, lr: 1.0e-04 loss: 1.020e-01]
anha:  67%|######6   | 2202/3300 [07:53<33:29,  1.83s/it, lr: 1.0e-04 loss: 1.327e-01]
anha:  67%|######6   | 2202/3300 [07:53<33:29,  1.83s/it, lr: 1.0e-04 loss: 1.327e-01]
anha:  67%|######6   | 2203/3300 [07:53<34:10,  1.87s/it, lr: 1.0e-04 loss: 1.327e-01]
anha:  67%|######6   | 2203/3300 [07:53<34:10,  1.87s/it, lr: 1.0e-04 loss: 1.327e-01]
anha:  67%|######6   | 2203/3300 [07:55<34:10,  1.87s/it, lr: 1.0e-04 loss: 1.533e-01]
anha:  67%|######6   | 2203/3300 [07:55<34:10,  1.87s/it, lr: 1.0e-04 loss: 1.533e-01]
anha:  67%|######6   | 2204/3300 [07:55<33:27,  1.83s/it, lr: 1.0e-04 loss: 1.533e-01]
anha:  67%|######6   | 2204/3300 [07:55<33:27,  1.83s/it, lr: 1.0e-04 loss: 1.533e-01]
anha:  67%|######6   | 2204/3300 [07:56<33:27,  1.83s/it, lr: 1.0e-04 loss: 1.069e-01]
anha:  67%|######6   | 2204/3300 [07:56<33:27,  1.83s/it, lr: 1.0e-04 loss: 1.069e-01]
anha:  67%|######6   | 2205/3300 [07:57<32:55,  1.80s/it, lr: 1.0e-04 loss: 1.069e-01]
anha:  67%|######6   | 2205/3300 [07:57<32:55,  1.80s/it, lr: 1.0e-04 loss: 1.069e-01]
anha:  67%|######6   | 2205/3300 [07:58<32:55,  1.80s/it, lr: 1.0e-04 loss: 4.666e-02]
anha:  67%|######6   | 2205/3300 [07:58<32:55,  1.80s/it, lr: 1.0e-04 loss: 4.666e-02]
anha:  67%|######6   | 2206/3300 [07:58<29:20,  1.61s/it, lr: 1.0e-04 loss: 4.666e-02]
anha:  67%|######6   | 2206/3300 [07:58<29:20,  1.61s/it, lr: 1.0e-04 loss: 4.666e-02]
anha:  67%|######6   | 2206/3300 [08:01<29:20,  1.61s/it, lr: 1.0e-04 loss: 6.107e-02]
anha:  67%|######6   | 2206/3300 [08:01<29:20,  1.61s/it, lr: 1.0e-04 loss: 6.107e-02]
anha:  67%|######6   | 2207/3300 [08:01<38:36,  2.12s/it, lr: 1.0e-04 loss: 6.107e-02]
anha:  67%|######6   | 2207/3300 [08:01<38:36,  2.12s/it, lr: 1.0e-04 loss: 6.107e-02]
anha:  67%|######6   | 2207/3300 [08:02<38:36,  2.12s/it, lr: 1.0e-04 loss: 1.015e-01]
anha:  67%|######6   | 2207/3300 [08:02<38:36,  2.12s/it, lr: 1.0e-04 loss: 1.015e-01]
anha:  67%|######6   | 2208/3300 [08:02<32:40,  1.80s/it, lr: 1.0e-04 loss: 1.015e-01]
anha:  67%|######6   | 2208/3300 [08:02<32:40,  1.80s/it, lr: 1.0e-04 loss: 1.015e-01]
anha:  67%|######6   | 2208/3300 [08:04<32:40,  1.80s/it, lr: 1.0e-04 loss: 5.083e-02]
anha:  67%|######6   | 2208/3300 [08:04<32:40,  1.80s/it, lr: 1.0e-04 loss: 5.083e-02]
anha:  67%|######6   | 2209/3300 [08:04<33:29,  1.84s/it, lr: 1.0e-04 loss: 5.083e-02]
anha:  67%|######6   | 2209/3300 [08:04<33:29,  1.84s/it, lr: 1.0e-04 loss: 5.083e-02]
anha:  67%|######6   | 2209/3300 [08:05<33:29,  1.84s/it, lr: 1.0e-04 loss: 5.775e-02]
anha:  67%|######6   | 2209/3300 [08:05<33:29,  1.84s/it, lr: 1.0e-04 loss: 5.775e-02]
anha:  67%|######6   | 2210/3300 [08:07<33:27,  1.84s/it, lr: 1.0e-04 loss: 6.022e-02]
anha:  67%|######6   | 2210/3300 [08:07<33:27,  1.84s/it, lr: 1.0e-04 loss: 6.022e-02]
anha:  67%|######7   | 2211/3300 [08:07<26:17,  1.45s/it, lr: 1.0e-04 loss: 6.022e-02]
anha:  67%|######7   | 2211/3300 [08:07<26:17,  1.45s/it, lr: 1.0e-04 loss: 6.022e-02]
anha:  67%|######7   | 2211/3300 [08:08<26:17,  1.45s/it, lr: 1.0e-04 loss: 1.104e-01]
anha:  67%|######7   | 2211/3300 [08:08<26:17,  1.45s/it, lr: 1.0e-04 loss: 1.104e-01]
anha:  67%|######7   | 2212/3300 [08:08<24:14,  1.34s/it, lr: 1.0e-04 loss: 1.104e-01]
anha:  67%|######7   | 2212/3300 [08:08<24:14,  1.34s/it, lr: 1.0e-04 loss: 1.104e-01]
anha:  67%|######7   | 2212/3300 [08:10<24:14,  1.34s/it, lr: 1.0e-04 loss: 7.972e-02]
anha:  67%|######7   | 2212/3300 [08:10<24:14,  1.34s/it, lr: 1.0e-04 loss: 7.972e-02]
anha:  67%|######7   | 2213/3300 [08:10<27:07,  1.50s/it, lr: 1.0e-04 loss: 7.972e-02]
anha:  67%|######7   | 2213/3300 [08:10<27:07,  1.50s/it, lr: 1.0e-04 loss: 7.972e-02]
anha:  67%|######7   | 2213/3300 [08:11<27:07,  1.50s/it, lr: 1.0e-04 loss: 6.255e-02]
anha:  67%|######7   | 2213/3300 [08:11<27:07,  1.50s/it, lr: 1.0e-04 loss: 6.255e-02]
anha:  67%|######7   | 2214/3300 [08:11<24:42,  1.37s/it, lr: 1.0e-04 loss: 6.255e-02]
anha:  67%|######7   | 2214/3300 [08:11<24:42,  1.37s/it, lr: 1.0e-04 loss: 6.255e-02]
anha:  67%|######7   | 2214/3300 [08:14<24:42,  1.37s/it, lr: 1.0e-04 loss: 5.950e-02]
anha:  67%|######7   | 2214/3300 [08:14<24:42,  1.37s/it, lr: 1.0e-04 loss: 5.950e-02]
anha:  67%|######7   | 2215/3300 [08:14<35:29,  1.96s/it, lr: 1.0e-04 loss: 5.950e-02]
anha:  67%|######7   | 2215/3300 [08:14<35:29,  1.96s/it, lr: 1.0e-04 loss: 5.950e-02]
anha:  67%|######7   | 2215/3300 [08:18<35:29,  1.96s/it, lr: 1.0e-04 loss: 2.761e-02]
anha:  67%|######7   | 2215/3300 [08:18<35:29,  1.96s/it, lr: 1.0e-04 loss: 2.761e-02]
anha:  67%|######7   | 2216/3300 [08:18<42:49,  2.37s/it, lr: 1.0e-04 loss: 2.761e-02]
anha:  67%|######7   | 2216/3300 [08:18<42:49,  2.37s/it, lr: 1.0e-04 loss: 2.761e-02]
anha:  67%|######7   | 2216/3300 [08:21<42:49,  2.37s/it, lr: 1.0e-04 loss: 4.771e-02]
anha:  67%|######7   | 2216/3300 [08:21<42:49,  2.37s/it, lr: 1.0e-04 loss: 4.771e-02]
anha:  67%|######7   | 2217/3300 [08:21<48:35,  2.69s/it, lr: 1.0e-04 loss: 4.771e-02]
anha:  67%|######7   | 2217/3300 [08:21<48:35,  2.69s/it, lr: 1.0e-04 loss: 4.771e-02]
anha:  67%|######7   | 2217/3300 [08:22<48:35,  2.69s/it, lr: 1.0e-04 loss: 9.042e-02]
anha:  67%|######7   | 2217/3300 [08:22<48:35,  2.69s/it, lr: 1.0e-04 loss: 9.042e-02]
anha:  67%|######7   | 2218/3300 [08:22<39:40,  2.20s/it, lr: 1.0e-04 loss: 9.042e-02]
anha:  67%|######7   | 2218/3300 [08:22<39:40,  2.20s/it, lr: 1.0e-04 loss: 9.042e-02]
anha:  67%|######7   | 2218/3300 [08:26<39:40,  2.20s/it, lr: 1.0e-04 loss: 9.120e-02]
anha:  67%|######7   | 2218/3300 [08:26<39:40,  2.20s/it, lr: 1.0e-04 loss: 9.120e-02]
anha:  67%|######7   | 2219/3300 [08:26<46:22,  2.57s/it, lr: 1.0e-04 loss: 9.120e-02]
anha:  67%|######7   | 2219/3300 [08:26<46:22,  2.57s/it, lr: 1.0e-04 loss: 9.120e-02]
anha:  67%|######7   | 2219/3300 [08:27<46:22,  2.57s/it, lr: 1.0e-04 loss: 3.435e-02]
anha:  67%|######7   | 2219/3300 [08:27<46:22,  2.57s/it, lr: 1.0e-04 loss: 3.435e-02]
anha:  67%|######7   | 2220/3300 [08:28<46:19,  2.57s/it, lr: 1.0e-04 loss: 8.300e-02]
anha:  67%|######7   | 2220/3300 [08:28<46:19,  2.57s/it, lr: 1.0e-04 loss: 8.300e-02]
anha:  67%|######7   | 2221/3300 [08:28<29:15,  1.63s/it, lr: 1.0e-04 loss: 8.300e-02]
anha:  67%|######7   | 2221/3300 [08:28<29:15,  1.63s/it, lr: 1.0e-04 loss: 8.300e-02]
anha:  67%|######7   | 2221/3300 [08:32<29:15,  1.63s/it, lr: 1.0e-04 loss: 4.341e-02]
anha:  67%|######7   | 2221/3300 [08:32<29:15,  1.63s/it, lr: 1.0e-04 loss: 4.341e-02]
anha:  67%|######7   | 2222/3300 [08:32<36:20,  2.02s/it, lr: 1.0e-04 loss: 4.341e-02]
anha:  67%|######7   | 2222/3300 [08:32<36:20,  2.02s/it, lr: 1.0e-04 loss: 4.341e-02]
anha:  67%|######7   | 2222/3300 [08:33<36:20,  2.02s/it, lr: 1.0e-04 loss: 6.721e-02]
anha:  67%|######7   | 2222/3300 [08:33<36:20,  2.02s/it, lr: 1.0e-04 loss: 6.721e-02]
anha:  67%|######7   | 2223/3300 [08:33<31:35,  1.76s/it, lr: 1.0e-04 loss: 6.721e-02]
anha:  67%|######7   | 2223/3300 [08:33<31:35,  1.76s/it, lr: 1.0e-04 loss: 6.721e-02]
anha:  67%|######7   | 2223/3300 [08:34<31:35,  1.76s/it, lr: 1.0e-04 loss: 5.013e-02]
anha:  67%|######7   | 2223/3300 [08:34<31:35,  1.76s/it, lr: 1.0e-04 loss: 5.013e-02]
anha:  67%|######7   | 2224/3300 [08:34<28:26,  1.59s/it, lr: 1.0e-04 loss: 5.013e-02]
anha:  67%|######7   | 2224/3300 [08:34<28:26,  1.59s/it, lr: 1.0e-04 loss: 5.013e-02]
anha:  67%|######7   | 2224/3300 [08:36<28:26,  1.59s/it, lr: 1.0e-04 loss: 6.921e-02]
anha:  67%|######7   | 2224/3300 [08:36<28:26,  1.59s/it, lr: 1.0e-04 loss: 6.921e-02]
anha:  67%|######7   | 2225/3300 [08:36<30:21,  1.69s/it, lr: 1.0e-04 loss: 6.921e-02]
anha:  67%|######7   | 2225/3300 [08:36<30:21,  1.69s/it, lr: 1.0e-04 loss: 6.921e-02]
anha:  67%|######7   | 2225/3300 [08:37<30:21,  1.69s/it, lr: 1.0e-04 loss: 6.692e-02]
anha:  67%|######7   | 2225/3300 [08:37<30:21,  1.69s/it, lr: 1.0e-04 loss: 6.692e-02]
anha:  67%|######7   | 2226/3300 [08:37<26:48,  1.50s/it, lr: 1.0e-04 loss: 6.692e-02]
anha:  67%|######7   | 2226/3300 [08:37<26:48,  1.50s/it, lr: 1.0e-04 loss: 6.692e-02]
anha:  67%|######7   | 2226/3300 [08:39<26:48,  1.50s/it, lr: 1.0e-04 loss: 1.305e-01]
anha:  67%|######7   | 2226/3300 [08:39<26:48,  1.50s/it, lr: 1.0e-04 loss: 1.305e-01]
anha:  67%|######7   | 2227/3300 [08:39<27:59,  1.57s/it, lr: 1.0e-04 loss: 1.305e-01]
anha:  67%|######7   | 2227/3300 [08:39<27:59,  1.57s/it, lr: 1.0e-04 loss: 1.305e-01]
anha:  67%|######7   | 2227/3300 [08:41<27:59,  1.57s/it, lr: 1.0e-04 loss: 5.193e-02]
anha:  67%|######7   | 2227/3300 [08:41<27:59,  1.57s/it, lr: 1.0e-04 loss: 5.193e-02]
anha:  68%|######7   | 2228/3300 [08:41<29:59,  1.68s/it, lr: 1.0e-04 loss: 5.193e-02]
anha:  68%|######7   | 2228/3300 [08:41<29:59,  1.68s/it, lr: 1.0e-04 loss: 5.193e-02]
anha:  68%|######7   | 2228/3300 [08:44<29:59,  1.68s/it, lr: 1.0e-04 loss: 1.062e-01]
anha:  68%|######7   | 2228/3300 [08:44<29:59,  1.68s/it, lr: 1.0e-04 loss: 1.062e-01]
anha:  68%|######7   | 2229/3300 [08:44<39:26,  2.21s/it, lr: 1.0e-04 loss: 1.062e-01]
anha:  68%|######7   | 2229/3300 [08:44<39:26,  2.21s/it, lr: 1.0e-04 loss: 1.062e-01]
anha:  68%|######7   | 2229/3300 [08:47<39:26,  2.21s/it, lr: 1.0e-04 loss: 5.222e-02]
anha:  68%|######7   | 2229/3300 [08:47<39:26,  2.21s/it, lr: 1.0e-04 loss: 5.222e-02]
anha:  68%|######7   | 2230/3300 [08:51<39:24,  2.21s/it, lr: 1.0e-04 loss: 3.408e-02]
anha:  68%|######7   | 2230/3300 [08:51<39:24,  2.21s/it, lr: 1.0e-04 loss: 3.408e-02]
anha:  68%|######7   | 2231/3300 [08:51<35:05,  1.97s/it, lr: 1.0e-04 loss: 3.408e-02]
anha:  68%|######7   | 2231/3300 [08:51<35:05,  1.97s/it, lr: 1.0e-04 loss: 3.408e-02]
anha:  68%|######7   | 2231/3300 [08:52<35:05,  1.97s/it, lr: 1.0e-04 loss: 9.991e-02]
anha:  68%|######7   | 2231/3300 [08:52<35:05,  1.97s/it, lr: 1.0e-04 loss: 9.991e-02]
anha:  68%|######7   | 2232/3300 [08:52<33:57,  1.91s/it, lr: 1.0e-04 loss: 9.991e-02]
anha:  68%|######7   | 2232/3300 [08:52<33:57,  1.91s/it, lr: 1.0e-04 loss: 9.991e-02]
anha:  68%|######7   | 2232/3300 [08:53<33:57,  1.91s/it, lr: 1.0e-04 loss: 7.319e-02]
anha:  68%|######7   | 2232/3300 [08:53<33:57,  1.91s/it, lr: 1.0e-04 loss: 7.319e-02]
anha:  68%|######7   | 2233/3300 [08:53<30:17,  1.70s/it, lr: 1.0e-04 loss: 7.319e-02]
anha:  68%|######7   | 2233/3300 [08:53<30:17,  1.70s/it, lr: 1.0e-04 loss: 7.319e-02]
anha:  68%|######7   | 2233/3300 [08:54<30:17,  1.70s/it, lr: 1.0e-04 loss: 7.692e-02]
anha:  68%|######7   | 2233/3300 [08:54<30:17,  1.70s/it, lr: 1.0e-04 loss: 7.692e-02]
anha:  68%|######7   | 2234/3300 [08:54<26:51,  1.51s/it, lr: 1.0e-04 loss: 7.692e-02]
anha:  68%|######7   | 2234/3300 [08:54<26:51,  1.51s/it, lr: 1.0e-04 loss: 7.692e-02]
anha:  68%|######7   | 2234/3300 [08:55<26:51,  1.51s/it, lr: 1.0e-04 loss: 4.689e-02]
anha:  68%|######7   | 2234/3300 [08:55<26:51,  1.51s/it, lr: 1.0e-04 loss: 4.689e-02]
anha:  68%|######7   | 2235/3300 [08:55<24:21,  1.37s/it, lr: 1.0e-04 loss: 4.689e-02]
anha:  68%|######7   | 2235/3300 [08:55<24:21,  1.37s/it, lr: 1.0e-04 loss: 4.689e-02]
anha:  68%|######7   | 2235/3300 [08:59<24:21,  1.37s/it, lr: 1.0e-04 loss: 7.599e-02]
anha:  68%|######7   | 2235/3300 [08:59<24:21,  1.37s/it, lr: 1.0e-04 loss: 7.599e-02]
anha:  68%|######7   | 2236/3300 [08:59<33:45,  1.90s/it, lr: 1.0e-04 loss: 7.599e-02]
anha:  68%|######7   | 2236/3300 [08:59<33:45,  1.90s/it, lr: 1.0e-04 loss: 7.599e-02]
anha:  68%|######7   | 2236/3300 [09:00<33:45,  1.90s/it, lr: 1.0e-04 loss: 5.057e-02]
anha:  68%|######7   | 2236/3300 [09:00<33:45,  1.90s/it, lr: 1.0e-04 loss: 5.057e-02]
anha:  68%|######7   | 2237/3300 [09:00<29:06,  1.64s/it, lr: 1.0e-04 loss: 5.057e-02]
anha:  68%|######7   | 2237/3300 [09:00<29:06,  1.64s/it, lr: 1.0e-04 loss: 5.057e-02]
anha:  68%|######7   | 2237/3300 [09:01<29:06,  1.64s/it, lr: 1.0e-04 loss: 5.879e-02]
anha:  68%|######7   | 2237/3300 [09:01<29:06,  1.64s/it, lr: 1.0e-04 loss: 5.879e-02]
anha:  68%|######7   | 2238/3300 [09:01<25:46,  1.46s/it, lr: 1.0e-04 loss: 5.879e-02]
anha:  68%|######7   | 2238/3300 [09:01<25:46,  1.46s/it, lr: 1.0e-04 loss: 5.879e-02]
anha:  68%|######7   | 2238/3300 [09:02<25:46,  1.46s/it, lr: 1.0e-04 loss: 6.464e-02]
anha:  68%|######7   | 2238/3300 [09:02<25:46,  1.46s/it, lr: 1.0e-04 loss: 6.464e-02]
anha:  68%|######7   | 2239/3300 [09:02<23:18,  1.32s/it, lr: 1.0e-04 loss: 6.464e-02]
anha:  68%|######7   | 2239/3300 [09:02<23:18,  1.32s/it, lr: 1.0e-04 loss: 6.464e-02]
anha:  68%|######7   | 2239/3300 [09:03<23:18,  1.32s/it, lr: 1.0e-04 loss: 6.328e-02]
anha:  68%|######7   | 2239/3300 [09:03<23:18,  1.32s/it, lr: 1.0e-04 loss: 6.328e-02]
anha:  68%|######7   | 2240/3300 [09:05<23:17,  1.32s/it, lr: 1.0e-04 loss: 7.378e-02]
anha:  68%|######7   | 2240/3300 [09:05<23:17,  1.32s/it, lr: 1.0e-04 loss: 7.378e-02]
anha:  68%|######7   | 2241/3300 [09:05<20:29,  1.16s/it, lr: 1.0e-04 loss: 7.378e-02]
anha:  68%|######7   | 2241/3300 [09:05<20:29,  1.16s/it, lr: 1.0e-04 loss: 7.378e-02]
anha:  68%|######7   | 2241/3300 [09:08<20:29,  1.16s/it, lr: 1.0e-04 loss: 3.219e-02]
anha:  68%|######7   | 2241/3300 [09:08<20:29,  1.16s/it, lr: 1.0e-04 loss: 3.219e-02]
anha:  68%|######7   | 2242/3300 [09:08<30:36,  1.74s/it, lr: 1.0e-04 loss: 3.219e-02]
anha:  68%|######7   | 2242/3300 [09:08<30:36,  1.74s/it, lr: 1.0e-04 loss: 3.219e-02]
anha:  68%|######7   | 2242/3300 [09:11<30:36,  1.74s/it, lr: 1.0e-04 loss: 6.859e-02]
anha:  68%|######7   | 2242/3300 [09:11<30:36,  1.74s/it, lr: 1.0e-04 loss: 6.859e-02]
anha:  68%|######7   | 2243/3300 [09:11<37:41,  2.14s/it, lr: 1.0e-04 loss: 6.859e-02]
anha:  68%|######7   | 2243/3300 [09:11<37:41,  2.14s/it, lr: 1.0e-04 loss: 6.859e-02]
anha:  68%|######7   | 2243/3300 [09:13<37:41,  2.14s/it, lr: 1.0e-04 loss: 1.750e-01]
anha:  68%|######7   | 2243/3300 [09:13<37:41,  2.14s/it, lr: 1.0e-04 loss: 1.750e-01]
anha:  68%|######8   | 2244/3300 [09:13<36:52,  2.10s/it, lr: 1.0e-04 loss: 1.750e-01]
anha:  68%|######8   | 2244/3300 [09:13<36:52,  2.10s/it, lr: 1.0e-04 loss: 1.750e-01]
anha:  68%|######8   | 2244/3300 [09:15<36:52,  2.10s/it, lr: 1.0e-04 loss: 7.561e-02]
anha:  68%|######8   | 2244/3300 [09:15<36:52,  2.10s/it, lr: 1.0e-04 loss: 7.561e-02]
anha:  68%|######8   | 2245/3300 [09:15<36:15,  2.06s/it, lr: 1.0e-04 loss: 7.561e-02]
anha:  68%|######8   | 2245/3300 [09:15<36:15,  2.06s/it, lr: 1.0e-04 loss: 7.561e-02]
anha:  68%|######8   | 2245/3300 [09:16<36:15,  2.06s/it, lr: 1.0e-04 loss: 1.338e-01]
anha:  68%|######8   | 2245/3300 [09:16<36:15,  2.06s/it, lr: 1.0e-04 loss: 1.338e-01]
anha:  68%|######8   | 2246/3300 [09:16<30:56,  1.76s/it, lr: 1.0e-04 loss: 1.338e-01]
anha:  68%|######8   | 2246/3300 [09:16<30:56,  1.76s/it, lr: 1.0e-04 loss: 1.338e-01]
anha:  68%|######8   | 2246/3300 [09:18<30:56,  1.76s/it, lr: 1.0e-04 loss: 1.984e-01]
anha:  68%|######8   | 2246/3300 [09:18<30:56,  1.76s/it, lr: 1.0e-04 loss: 1.984e-01]
anha:  68%|######8   | 2247/3300 [09:18<32:00,  1.82s/it, lr: 1.0e-04 loss: 1.984e-01]
anha:  68%|######8   | 2247/3300 [09:18<32:00,  1.82s/it, lr: 1.0e-04 loss: 1.984e-01]
anha:  68%|######8   | 2247/3300 [09:20<32:00,  1.82s/it, lr: 1.0e-04 loss: 4.679e-02]
anha:  68%|######8   | 2247/3300 [09:20<32:00,  1.82s/it, lr: 1.0e-04 loss: 4.679e-02]
anha:  68%|######8   | 2248/3300 [09:20<32:42,  1.87s/it, lr: 1.0e-04 loss: 4.679e-02]
anha:  68%|######8   | 2248/3300 [09:20<32:42,  1.87s/it, lr: 1.0e-04 loss: 4.679e-02]
anha:  68%|######8   | 2248/3300 [09:22<32:42,  1.87s/it, lr: 1.0e-04 loss: 1.128e-01]
anha:  68%|######8   | 2248/3300 [09:22<32:42,  1.87s/it, lr: 1.0e-04 loss: 1.128e-01]
anha:  68%|######8   | 2249/3300 [09:22<33:07,  1.89s/it, lr: 1.0e-04 loss: 1.128e-01]
anha:  68%|######8   | 2249/3300 [09:22<33:07,  1.89s/it, lr: 1.0e-04 loss: 1.128e-01]
anha:  68%|######8   | 2249/3300 [09:26<33:07,  1.89s/it, lr: 1.0e-04 loss: 9.616e-02]
anha:  68%|######8   | 2249/3300 [09:26<33:07,  1.89s/it, lr: 1.0e-04 loss: 9.616e-02]


Generating Images: 0it [00:00, ?it/s]
Generating Images: 0it [00:00, ?it/s]


                                     
                                     
anha:  68%|######8   | 2250/3300 [09:27<33:05,  1.89s/it, lr: 1.0e-04 loss: 6.319e-02]
anha:  68%|######8   | 2250/3300 [09:27<33:05,  1.89s/it, lr: 1.0e-04 loss: 6.319e-02]
anha:  68%|######8   | 2251/3300 [09:27<21:58,  1.26s/it, lr: 1.0e-04 loss: 6.319e-02]
anha:  68%|######8   | 2251/3300 [09:27<21:58,  1.26s/it, lr: 1.0e-04 loss: 6.319e-02]
anha:  68%|######8   | 2251/3300 [09:30<21:58,  1.26s/it, lr: 1.0e-04 loss: 8.133e-02]
anha:  68%|######8   | 2251/3300 [09:30<21:58,  1.26s/it, lr: 1.0e-04 loss: 8.133e-02]
anha:  68%|######8   | 2252/3300 [09:30<31:20,  1.79s/it, lr: 1.0e-04 loss: 8.133e-02]
anha:  68%|######8   | 2252/3300 [09:30<31:20,  1.79s/it, lr: 1.0e-04 loss: 8.133e-02]
anha:  68%|######8   | 2252/3300 [09:32<31:20,  1.79s/it, lr: 1.0e-04 loss: 7.805e-02]
anha:  68%|######8   | 2252/3300 [09:32<31:20,  1.79s/it, lr: 1.0e-04 loss: 7.805e-02]
anha:  68%|######8   | 2253/3300 [09:32<31:05,  1.78s/it, lr: 1.0e-04 loss: 7.805e-02]
anha:  68%|######8   | 2253/3300 [09:32<31:05,  1.78s/it, lr: 1.0e-04 loss: 7.805e-02]
anha:  68%|######8   | 2253/3300 [09:33<31:05,  1.78s/it, lr: 1.0e-04 loss: 1.776e-01]
anha:  68%|######8   | 2253/3300 [09:33<31:05,  1.78s/it, lr: 1.0e-04 loss: 1.776e-01]
anha:  68%|######8   | 2254/3300 [09:33<27:20,  1.57s/it, lr: 1.0e-04 loss: 1.776e-01]
anha:  68%|######8   | 2254/3300 [09:33<27:20,  1.57s/it, lr: 1.0e-04 loss: 1.776e-01]
anha:  68%|######8   | 2254/3300 [09:35<27:20,  1.57s/it, lr: 1.0e-04 loss: 5.065e-02]
anha:  68%|######8   | 2254/3300 [09:35<27:20,  1.57s/it, lr: 1.0e-04 loss: 5.065e-02]
anha:  68%|######8   | 2255/3300 [09:35<29:18,  1.68s/it, lr: 1.0e-04 loss: 5.065e-02]
anha:  68%|######8   | 2255/3300 [09:35<29:18,  1.68s/it, lr: 1.0e-04 loss: 5.065e-02]
anha:  68%|######8   | 2255/3300 [09:38<29:18,  1.68s/it, lr: 1.0e-04 loss: 3.283e-02]
anha:  68%|######8   | 2255/3300 [09:38<29:18,  1.68s/it, lr: 1.0e-04 loss: 3.283e-02]
anha:  68%|######8   | 2256/3300 [09:38<37:04,  2.13s/it, lr: 1.0e-04 loss: 3.283e-02]
anha:  68%|######8   | 2256/3300 [09:38<37:04,  2.13s/it, lr: 1.0e-04 loss: 3.283e-02]
anha:  68%|######8   | 2256/3300 [09:39<37:04,  2.13s/it, lr: 1.0e-04 loss: 3.949e-02]
anha:  68%|######8   | 2256/3300 [09:39<37:04,  2.13s/it, lr: 1.0e-04 loss: 3.949e-02]
anha:  68%|######8   | 2257/3300 [09:39<31:22,  1.81s/it, lr: 1.0e-04 loss: 3.949e-02]
anha:  68%|######8   | 2257/3300 [09:39<31:22,  1.81s/it, lr: 1.0e-04 loss: 3.949e-02]
anha:  68%|######8   | 2257/3300 [09:42<31:22,  1.81s/it, lr: 1.0e-04 loss: 4.478e-02]
anha:  68%|######8   | 2257/3300 [09:42<31:22,  1.81s/it, lr: 1.0e-04 loss: 4.478e-02]
anha:  68%|######8   | 2258/3300 [09:42<34:55,  2.01s/it, lr: 1.0e-04 loss: 4.478e-02]
anha:  68%|######8   | 2258/3300 [09:42<34:55,  2.01s/it, lr: 1.0e-04 loss: 4.478e-02]
anha:  68%|######8   | 2258/3300 [09:44<34:55,  2.01s/it, lr: 1.0e-04 loss: 4.027e-02]
anha:  68%|######8   | 2258/3300 [09:44<34:55,  2.01s/it, lr: 1.0e-04 loss: 4.027e-02]
anha:  68%|######8   | 2259/3300 [09:44<34:34,  1.99s/it, lr: 1.0e-04 loss: 4.027e-02]
anha:  68%|######8   | 2259/3300 [09:44<34:34,  1.99s/it, lr: 1.0e-04 loss: 4.027e-02]
anha:  68%|######8   | 2259/3300 [09:47<34:34,  1.99s/it, lr: 1.0e-04 loss: 1.380e-01]
anha:  68%|######8   | 2259/3300 [09:47<34:34,  1.99s/it, lr: 1.0e-04 loss: 1.380e-01]
anha:  68%|######8   | 2260/3300 [09:49<34:32,  1.99s/it, lr: 1.0e-04 loss: 5.124e-02]
anha:  68%|######8   | 2260/3300 [09:49<34:32,  1.99s/it, lr: 1.0e-04 loss: 5.124e-02]
anha:  69%|######8   | 2261/3300 [09:49<26:33,  1.53s/it, lr: 1.0e-04 loss: 5.124e-02]
anha:  69%|######8   | 2261/3300 [09:49<26:33,  1.53s/it, lr: 1.0e-04 loss: 5.124e-02]
anha:  69%|######8   | 2261/3300 [09:52<26:33,  1.53s/it, lr: 1.0e-04 loss: 4.022e-02]
anha:  69%|######8   | 2261/3300 [09:52<26:33,  1.53s/it, lr: 1.0e-04 loss: 4.022e-02]
anha:  69%|######8   | 2262/3300 [09:52<34:20,  1.99s/it, lr: 1.0e-04 loss: 4.022e-02]
anha:  69%|######8   | 2262/3300 [09:52<34:20,  1.99s/it, lr: 1.0e-04 loss: 4.022e-02]
anha:  69%|######8   | 2262/3300 [09:53<34:20,  1.99s/it, lr: 1.0e-04 loss: 9.851e-02]
anha:  69%|######8   | 2262/3300 [09:53<34:20,  1.99s/it, lr: 1.0e-04 loss: 9.851e-02]
anha:  69%|######8   | 2263/3300 [09:53<29:50,  1.73s/it, lr: 1.0e-04 loss: 9.851e-02]
anha:  69%|######8   | 2263/3300 [09:53<29:50,  1.73s/it, lr: 1.0e-04 loss: 9.851e-02]
anha:  69%|######8   | 2263/3300 [09:54<29:50,  1.73s/it, lr: 1.0e-04 loss: 1.215e-01]
anha:  69%|######8   | 2263/3300 [09:54<29:50,  1.73s/it, lr: 1.0e-04 loss: 1.215e-01]
anha:  69%|######8   | 2264/3300 [09:54<26:27,  1.53s/it, lr: 1.0e-04 loss: 1.215e-01]
anha:  69%|######8   | 2264/3300 [09:54<26:27,  1.53s/it, lr: 1.0e-04 loss: 1.215e-01]
anha:  69%|######8   | 2264/3300 [09:58<26:27,  1.53s/it, lr: 1.0e-04 loss: 3.681e-02]
anha:  69%|######8   | 2264/3300 [09:58<26:27,  1.53s/it, lr: 1.0e-04 loss: 3.681e-02]
anha:  69%|######8   | 2265/3300 [09:58<35:48,  2.08s/it, lr: 1.0e-04 loss: 3.681e-02]
anha:  69%|######8   | 2265/3300 [09:58<35:48,  2.08s/it, lr: 1.0e-04 loss: 3.681e-02]
anha:  69%|######8   | 2265/3300 [10:00<35:48,  2.08s/it, lr: 1.0e-04 loss: 2.268e-01]
anha:  69%|######8   | 2265/3300 [10:00<35:48,  2.08s/it, lr: 1.0e-04 loss: 2.268e-01]
anha:  69%|######8   | 2266/3300 [10:00<35:15,  2.05s/it, lr: 1.0e-04 loss: 2.268e-01]
anha:  69%|######8   | 2266/3300 [10:00<35:15,  2.05s/it, lr: 1.0e-04 loss: 2.268e-01]
anha:  69%|######8   | 2266/3300 [10:02<35:15,  2.05s/it, lr: 1.0e-04 loss: 3.321e-02]
anha:  69%|######8   | 2266/3300 [10:02<35:15,  2.05s/it, lr: 1.0e-04 loss: 3.321e-02]
anha:  69%|######8   | 2267/3300 [10:02<34:53,  2.03s/it, lr: 1.0e-04 loss: 3.321e-02]
anha:  69%|######8   | 2267/3300 [10:02<34:53,  2.03s/it, lr: 1.0e-04 loss: 3.321e-02]
anha:  69%|######8   | 2267/3300 [10:05<34:53,  2.03s/it, lr: 1.0e-04 loss: 1.373e-01]
anha:  69%|######8   | 2267/3300 [10:05<34:53,  2.03s/it, lr: 1.0e-04 loss: 1.373e-01]
anha:  69%|######8   | 2268/3300 [10:05<41:02,  2.39s/it, lr: 1.0e-04 loss: 1.373e-01]
anha:  69%|######8   | 2268/3300 [10:05<41:02,  2.39s/it, lr: 1.0e-04 loss: 1.373e-01]
anha:  69%|######8   | 2268/3300 [10:07<41:02,  2.39s/it, lr: 1.0e-04 loss: 4.747e-02]
anha:  69%|######8   | 2268/3300 [10:07<41:02,  2.39s/it, lr: 1.0e-04 loss: 4.747e-02]
anha:  69%|######8   | 2269/3300 [10:07<38:06,  2.22s/it, lr: 1.0e-04 loss: 4.747e-02]
anha:  69%|######8   | 2269/3300 [10:07<38:06,  2.22s/it, lr: 1.0e-04 loss: 4.747e-02]
anha:  69%|######8   | 2269/3300 [10:09<38:06,  2.22s/it, lr: 1.0e-04 loss: 4.565e-02]
anha:  69%|######8   | 2269/3300 [10:09<38:06,  2.22s/it, lr: 1.0e-04 loss: 4.565e-02]
anha:  69%|######8   | 2270/3300 [10:10<38:04,  2.22s/it, lr: 1.0e-04 loss: 5.538e-02]
anha:  69%|######8   | 2270/3300 [10:10<38:04,  2.22s/it, lr: 1.0e-04 loss: 5.538e-02]
anha:  69%|######8   | 2271/3300 [10:10<24:34,  1.43s/it, lr: 1.0e-04 loss: 5.538e-02]
anha:  69%|######8   | 2271/3300 [10:10<24:34,  1.43s/it, lr: 1.0e-04 loss: 5.538e-02]
anha:  69%|######8   | 2271/3300 [10:12<24:34,  1.43s/it, lr: 1.0e-04 loss: 1.360e-01]
anha:  69%|######8   | 2271/3300 [10:12<24:34,  1.43s/it, lr: 1.0e-04 loss: 1.360e-01]
anha:  69%|######8   | 2272/3300 [10:12<26:47,  1.56s/it, lr: 1.0e-04 loss: 1.360e-01]
anha:  69%|######8   | 2272/3300 [10:12<26:47,  1.56s/it, lr: 1.0e-04 loss: 1.360e-01]
anha:  69%|######8   | 2272/3300 [10:14<26:47,  1.56s/it, lr: 1.0e-04 loss: 6.866e-02]
anha:  69%|######8   | 2272/3300 [10:14<26:47,  1.56s/it, lr: 1.0e-04 loss: 6.866e-02]
anha:  69%|######8   | 2273/3300 [10:14<28:34,  1.67s/it, lr: 1.0e-04 loss: 6.866e-02]
anha:  69%|######8   | 2273/3300 [10:14<28:34,  1.67s/it, lr: 1.0e-04 loss: 6.866e-02]
anha:  69%|######8   | 2273/3300 [10:17<28:34,  1.67s/it, lr: 1.0e-04 loss: 5.042e-02]
anha:  69%|######8   | 2273/3300 [10:17<28:34,  1.67s/it, lr: 1.0e-04 loss: 5.042e-02]
anha:  69%|######8   | 2274/3300 [10:17<36:54,  2.16s/it, lr: 1.0e-04 loss: 5.042e-02]
anha:  69%|######8   | 2274/3300 [10:17<36:54,  2.16s/it, lr: 1.0e-04 loss: 5.042e-02]
anha:  69%|######8   | 2274/3300 [10:18<36:54,  2.16s/it, lr: 1.0e-04 loss: 3.067e-02]
anha:  69%|######8   | 2274/3300 [10:18<36:54,  2.16s/it, lr: 1.0e-04 loss: 3.067e-02]
anha:  69%|######8   | 2275/3300 [10:18<31:21,  1.84s/it, lr: 1.0e-04 loss: 3.067e-02]
anha:  69%|######8   | 2275/3300 [10:18<31:21,  1.84s/it, lr: 1.0e-04 loss: 3.067e-02]
anha:  69%|######8   | 2275/3300 [10:22<31:21,  1.84s/it, lr: 1.0e-04 loss: 9.646e-02]
anha:  69%|######8   | 2275/3300 [10:22<31:21,  1.84s/it, lr: 1.0e-04 loss: 9.646e-02]
anha:  69%|######8   | 2276/3300 [10:22<38:23,  2.25s/it, lr: 1.0e-04 loss: 9.646e-02]
anha:  69%|######8   | 2276/3300 [10:22<38:23,  2.25s/it, lr: 1.0e-04 loss: 9.646e-02]
anha:  69%|######8   | 2276/3300 [10:23<38:23,  2.25s/it, lr: 1.0e-04 loss: 3.891e-02]
anha:  69%|######8   | 2276/3300 [10:23<38:23,  2.25s/it, lr: 1.0e-04 loss: 3.891e-02]
anha:  69%|######9   | 2277/3300 [10:23<35:43,  2.10s/it, lr: 1.0e-04 loss: 3.891e-02]
anha:  69%|######9   | 2277/3300 [10:23<35:43,  2.10s/it, lr: 1.0e-04 loss: 3.891e-02]
anha:  69%|######9   | 2277/3300 [10:24<35:43,  2.10s/it, lr: 1.0e-04 loss: 5.969e-02]
anha:  69%|######9   | 2277/3300 [10:24<35:43,  2.10s/it, lr: 1.0e-04 loss: 5.969e-02]
anha:  69%|######9   | 2278/3300 [10:24<30:15,  1.78s/it, lr: 1.0e-04 loss: 5.969e-02]
anha:  69%|######9   | 2278/3300 [10:24<30:15,  1.78s/it, lr: 1.0e-04 loss: 5.969e-02]
anha:  69%|######9   | 2278/3300 [10:25<30:15,  1.78s/it, lr: 1.0e-04 loss: 8.258e-02]
anha:  69%|######9   | 2278/3300 [10:25<30:15,  1.78s/it, lr: 1.0e-04 loss: 8.258e-02]
anha:  69%|######9   | 2279/3300 [10:25<26:51,  1.58s/it, lr: 1.0e-04 loss: 8.258e-02]
anha:  69%|######9   | 2279/3300 [10:25<26:51,  1.58s/it, lr: 1.0e-04 loss: 8.258e-02]
anha:  69%|######9   | 2279/3300 [10:26<26:51,  1.58s/it, lr: 1.0e-04 loss: 1.255e-01]
anha:  69%|######9   | 2279/3300 [10:26<26:51,  1.58s/it, lr: 1.0e-04 loss: 1.255e-01]
anha:  69%|######9   | 2280/3300 [10:27<26:50,  1.58s/it, lr: 1.0e-04 loss: 5.379e-02]
anha:  69%|######9   | 2280/3300 [10:27<26:50,  1.58s/it, lr: 1.0e-04 loss: 5.379e-02]
anha:  69%|######9   | 2281/3300 [10:27<18:24,  1.08s/it, lr: 1.0e-04 loss: 5.379e-02]
anha:  69%|######9   | 2281/3300 [10:27<18:24,  1.08s/it, lr: 1.0e-04 loss: 5.379e-02]
anha:  69%|######9   | 2281/3300 [10:31<18:24,  1.08s/it, lr: 1.0e-04 loss: 4.677e-02]
anha:  69%|######9   | 2281/3300 [10:31<18:24,  1.08s/it, lr: 1.0e-04 loss: 4.677e-02]
anha:  69%|######9   | 2282/3300 [10:31<28:24,  1.67s/it, lr: 1.0e-04 loss: 4.677e-02]
anha:  69%|######9   | 2282/3300 [10:31<28:24,  1.67s/it, lr: 1.0e-04 loss: 4.677e-02]
anha:  69%|######9   | 2282/3300 [10:32<28:24,  1.67s/it, lr: 1.0e-04 loss: 1.355e-01]
anha:  69%|######9   | 2282/3300 [10:32<28:24,  1.67s/it, lr: 1.0e-04 loss: 1.355e-01]
anha:  69%|######9   | 2283/3300 [10:32<25:27,  1.50s/it, lr: 1.0e-04 loss: 1.355e-01]
anha:  69%|######9   | 2283/3300 [10:32<25:27,  1.50s/it, lr: 1.0e-04 loss: 1.355e-01]
anha:  69%|######9   | 2283/3300 [10:33<25:27,  1.50s/it, lr: 1.0e-04 loss: 5.810e-02]
anha:  69%|######9   | 2283/3300 [10:33<25:27,  1.50s/it, lr: 1.0e-04 loss: 5.810e-02]
anha:  69%|######9   | 2284/3300 [10:33<23:05,  1.36s/it, lr: 1.0e-04 loss: 5.810e-02]
anha:  69%|######9   | 2284/3300 [10:33<23:05,  1.36s/it, lr: 1.0e-04 loss: 5.810e-02]
anha:  69%|######9   | 2284/3300 [10:35<23:05,  1.36s/it, lr: 1.0e-04 loss: 6.380e-01]
anha:  69%|######9   | 2284/3300 [10:35<23:05,  1.36s/it, lr: 1.0e-04 loss: 6.380e-01]
anha:  69%|######9   | 2285/3300 [10:35<24:47,  1.47s/it, lr: 1.0e-04 loss: 6.380e-01]
anha:  69%|######9   | 2285/3300 [10:35<24:47,  1.47s/it, lr: 1.0e-04 loss: 6.380e-01]
anha:  69%|######9   | 2285/3300 [10:37<24:47,  1.47s/it, lr: 1.0e-04 loss: 2.796e-02]
anha:  69%|######9   | 2285/3300 [10:37<24:47,  1.47s/it, lr: 1.0e-04 loss: 2.796e-02]
anha:  69%|######9   | 2286/3300 [10:37<27:05,  1.60s/it, lr: 1.0e-04 loss: 2.796e-02]
anha:  69%|######9   | 2286/3300 [10:37<27:05,  1.60s/it, lr: 1.0e-04 loss: 2.796e-02]
anha:  69%|######9   | 2286/3300 [10:40<27:05,  1.60s/it, lr: 1.0e-04 loss: 7.580e-02]
anha:  69%|######9   | 2286/3300 [10:40<27:05,  1.60s/it, lr: 1.0e-04 loss: 7.580e-02]
anha:  69%|######9   | 2287/3300 [10:40<34:44,  2.06s/it, lr: 1.0e-04 loss: 7.580e-02]
anha:  69%|######9   | 2287/3300 [10:40<34:44,  2.06s/it, lr: 1.0e-04 loss: 7.580e-02]
anha:  69%|######9   | 2287/3300 [10:43<34:44,  2.06s/it, lr: 1.0e-04 loss: 1.144e-01]
anha:  69%|######9   | 2287/3300 [10:43<34:44,  2.06s/it, lr: 1.0e-04 loss: 1.144e-01]
anha:  69%|######9   | 2288/3300 [10:43<41:21,  2.45s/it, lr: 1.0e-04 loss: 1.144e-01]
anha:  69%|######9   | 2288/3300 [10:43<41:21,  2.45s/it, lr: 1.0e-04 loss: 1.144e-01]
anha:  69%|######9   | 2288/3300 [10:44<41:21,  2.45s/it, lr: 1.0e-04 loss: 6.691e-02]
anha:  69%|######9   | 2288/3300 [10:44<41:21,  2.45s/it, lr: 1.0e-04 loss: 6.691e-02]
anha:  69%|######9   | 2289/3300 [10:44<34:15,  2.03s/it, lr: 1.0e-04 loss: 6.691e-02]
anha:  69%|######9   | 2289/3300 [10:44<34:15,  2.03s/it, lr: 1.0e-04 loss: 6.691e-02]
anha:  69%|######9   | 2289/3300 [10:48<34:15,  2.03s/it, lr: 1.0e-04 loss: 4.662e-02]
anha:  69%|######9   | 2289/3300 [10:48<34:15,  2.03s/it, lr: 1.0e-04 loss: 4.662e-02]
anha:  69%|######9   | 2290/3300 [10:49<34:13,  2.03s/it, lr: 1.0e-04 loss: 2.218e-01]
anha:  69%|######9   | 2290/3300 [10:49<34:13,  2.03s/it, lr: 1.0e-04 loss: 2.218e-01]
anha:  69%|######9   | 2291/3300 [10:49<22:25,  1.33s/it, lr: 1.0e-04 loss: 2.218e-01]
anha:  69%|######9   | 2291/3300 [10:49<22:25,  1.33s/it, lr: 1.0e-04 loss: 2.218e-01]
anha:  69%|######9   | 2291/3300 [10:52<22:25,  1.33s/it, lr: 1.0e-04 loss: 3.138e-01]
anha:  69%|######9   | 2291/3300 [10:52<22:25,  1.33s/it, lr: 1.0e-04 loss: 3.138e-01]
anha:  69%|######9   | 2292/3300 [10:52<30:49,  1.84s/it, lr: 1.0e-04 loss: 3.138e-01]
anha:  69%|######9   | 2292/3300 [10:52<30:49,  1.84s/it, lr: 1.0e-04 loss: 3.138e-01]
anha:  69%|######9   | 2292/3300 [10:55<30:49,  1.84s/it, lr: 1.0e-04 loss: 7.970e-02]
anha:  69%|######9   | 2292/3300 [10:55<30:49,  1.84s/it, lr: 1.0e-04 loss: 7.970e-02]
anha:  69%|######9   | 2293/3300 [10:55<37:08,  2.21s/it, lr: 1.0e-04 loss: 7.970e-02]
anha:  69%|######9   | 2293/3300 [10:55<37:08,  2.21s/it, lr: 1.0e-04 loss: 7.970e-02]
anha:  69%|######9   | 2293/3300 [10:57<37:08,  2.21s/it, lr: 1.0e-04 loss: 1.155e-01]
anha:  69%|######9   | 2293/3300 [10:57<37:08,  2.21s/it, lr: 1.0e-04 loss: 1.155e-01]
anha:  70%|######9   | 2294/3300 [10:57<34:54,  2.08s/it, lr: 1.0e-04 loss: 1.155e-01]
anha:  70%|######9   | 2294/3300 [10:57<34:54,  2.08s/it, lr: 1.0e-04 loss: 1.155e-01]
anha:  70%|######9   | 2294/3300 [10:58<34:54,  2.08s/it, lr: 1.0e-04 loss: 5.343e-02]
anha:  70%|######9   | 2294/3300 [10:58<34:54,  2.08s/it, lr: 1.0e-04 loss: 5.343e-02]
anha:  70%|######9   | 2295/3300 [10:58<29:51,  1.78s/it, lr: 1.0e-04 loss: 5.343e-02]
anha:  70%|######9   | 2295/3300 [10:58<29:51,  1.78s/it, lr: 1.0e-04 loss: 5.343e-02]
anha:  70%|######9   | 2295/3300 [11:00<29:51,  1.78s/it, lr: 1.0e-04 loss: 5.673e-02]
anha:  70%|######9   | 2295/3300 [11:00<29:51,  1.78s/it, lr: 1.0e-04 loss: 5.673e-02]
anha:  70%|######9   | 2296/3300 [11:00<29:34,  1.77s/it, lr: 1.0e-04 loss: 5.673e-02]
anha:  70%|######9   | 2296/3300 [11:00<29:34,  1.77s/it, lr: 1.0e-04 loss: 5.673e-02]
anha:  70%|######9   | 2296/3300 [11:01<29:34,  1.77s/it, lr: 1.0e-04 loss: 3.477e-01]
anha:  70%|######9   | 2296/3300 [11:01<29:34,  1.77s/it, lr: 1.0e-04 loss: 3.477e-01]
anha:  70%|######9   | 2297/3300 [11:01<26:20,  1.58s/it, lr: 1.0e-04 loss: 3.477e-01]
anha:  70%|######9   | 2297/3300 [11:01<26:20,  1.58s/it, lr: 1.0e-04 loss: 3.477e-01]
anha:  70%|######9   | 2297/3300 [11:04<26:20,  1.58s/it, lr: 1.0e-04 loss: 5.702e-02]
anha:  70%|######9   | 2297/3300 [11:04<26:20,  1.58s/it, lr: 1.0e-04 loss: 5.702e-02]
anha:  70%|######9   | 2298/3300 [11:04<35:25,  2.12s/it, lr: 1.0e-04 loss: 5.702e-02]
anha:  70%|######9   | 2298/3300 [11:04<35:25,  2.12s/it, lr: 1.0e-04 loss: 5.702e-02]
anha:  70%|######9   | 2298/3300 [11:05<35:25,  2.12s/it, lr: 1.0e-04 loss: 6.619e-02]
anha:  70%|######9   | 2298/3300 [11:05<35:25,  2.12s/it, lr: 1.0e-04 loss: 6.619e-02]
anha:  70%|######9   | 2299/3300 [11:05<29:51,  1.79s/it, lr: 1.0e-04 loss: 6.619e-02]
anha:  70%|######9   | 2299/3300 [11:05<29:51,  1.79s/it, lr: 1.0e-04 loss: 6.619e-02]
anha:  70%|######9   | 2299/3300 [11:09<29:51,  1.79s/it, lr: 1.0e-04 loss: 6.490e-02]
anha:  70%|######9   | 2299/3300 [11:09<29:51,  1.79s/it, lr: 1.0e-04 loss: 6.490e-02]
anha:  70%|######9   | 2300/3300 [11:10<29:49,  1.79s/it, lr: 1.0e-04 loss: 8.231e-02]
anha:  70%|######9   | 2300/3300 [11:10<29:49,  1.79s/it, lr: 1.0e-04 loss: 8.231e-02]
anha:  70%|######9   | 2301/3300 [11:10<20:04,  1.21s/it, lr: 1.0e-04 loss: 8.231e-02]
anha:  70%|######9   | 2301/3300 [11:10<20:04,  1.21s/it, lr: 1.0e-04 loss: 8.231e-02]
anha:  70%|######9   | 2301/3300 [11:11<20:04,  1.21s/it, lr: 1.0e-04 loss: 1.168e-01]
anha:  70%|######9   | 2301/3300 [11:11<20:04,  1.21s/it, lr: 1.0e-04 loss: 1.168e-01]
anha:  70%|######9   | 2302/3300 [11:11<19:15,  1.16s/it, lr: 1.0e-04 loss: 1.168e-01]
anha:  70%|######9   | 2302/3300 [11:11<19:15,  1.16s/it, lr: 1.0e-04 loss: 1.168e-01]
anha:  70%|######9   | 2302/3300 [11:12<19:15,  1.16s/it, lr: 1.0e-04 loss: 1.381e-01]
anha:  70%|######9   | 2302/3300 [11:12<19:15,  1.16s/it, lr: 1.0e-04 loss: 1.381e-01]
anha:  70%|######9   | 2303/3300 [11:12<21:39,  1.30s/it, lr: 1.0e-04 loss: 1.381e-01]
anha:  70%|######9   | 2303/3300 [11:12<21:39,  1.30s/it, lr: 1.0e-04 loss: 1.381e-01]
anha:  70%|######9   | 2303/3300 [11:16<21:39,  1.30s/it, lr: 1.0e-04 loss: 5.034e-02]
anha:  70%|######9   | 2303/3300 [11:16<21:39,  1.30s/it, lr: 1.0e-04 loss: 5.034e-02]
anha:  70%|######9   | 2304/3300 [11:16<31:25,  1.89s/it, lr: 1.0e-04 loss: 5.034e-02]
anha:  70%|######9   | 2304/3300 [11:16<31:25,  1.89s/it, lr: 1.0e-04 loss: 5.034e-02]
anha:  70%|######9   | 2304/3300 [11:17<31:25,  1.89s/it, lr: 1.0e-04 loss: 5.643e-02]
anha:  70%|######9   | 2304/3300 [11:17<31:25,  1.89s/it, lr: 1.0e-04 loss: 5.643e-02]
anha:  70%|######9   | 2305/3300 [11:17<27:17,  1.65s/it, lr: 1.0e-04 loss: 5.643e-02]
anha:  70%|######9   | 2305/3300 [11:17<27:17,  1.65s/it, lr: 1.0e-04 loss: 5.643e-02]
anha:  70%|######9   | 2305/3300 [11:20<27:17,  1.65s/it, lr: 1.0e-04 loss: 1.006e-01]
anha:  70%|######9   | 2305/3300 [11:20<27:17,  1.65s/it, lr: 1.0e-04 loss: 1.006e-01]
anha:  70%|######9   | 2306/3300 [11:20<35:00,  2.11s/it, lr: 1.0e-04 loss: 1.006e-01]
anha:  70%|######9   | 2306/3300 [11:20<35:00,  2.11s/it, lr: 1.0e-04 loss: 1.006e-01]
anha:  70%|######9   | 2306/3300 [11:24<35:00,  2.11s/it, lr: 1.0e-04 loss: 3.663e-02]
anha:  70%|######9   | 2306/3300 [11:24<35:00,  2.11s/it, lr: 1.0e-04 loss: 3.663e-02]
anha:  70%|######9   | 2307/3300 [11:24<41:15,  2.49s/it, lr: 1.0e-04 loss: 3.663e-02]
anha:  70%|######9   | 2307/3300 [11:24<41:15,  2.49s/it, lr: 1.0e-04 loss: 3.663e-02]
anha:  70%|######9   | 2307/3300 [11:26<41:15,  2.49s/it, lr: 1.0e-04 loss: 3.828e-01]
anha:  70%|######9   | 2307/3300 [11:26<41:15,  2.49s/it, lr: 1.0e-04 loss: 3.828e-01]
anha:  70%|######9   | 2308/3300 [11:26<38:35,  2.33s/it, lr: 1.0e-04 loss: 3.828e-01]
anha:  70%|######9   | 2308/3300 [11:26<38:35,  2.33s/it, lr: 1.0e-04 loss: 3.828e-01]
anha:  70%|######9   | 2308/3300 [11:29<38:35,  2.33s/it, lr: 1.0e-04 loss: 2.894e-02]
anha:  70%|######9   | 2308/3300 [11:29<38:35,  2.33s/it, lr: 1.0e-04 loss: 2.894e-02]
anha:  70%|######9   | 2309/3300 [11:29<43:41,  2.65s/it, lr: 1.0e-04 loss: 2.894e-02]
anha:  70%|######9   | 2309/3300 [11:29<43:41,  2.65s/it, lr: 1.0e-04 loss: 2.894e-02]
anha:  70%|######9   | 2309/3300 [11:31<43:41,  2.65s/it, lr: 1.0e-04 loss: 5.998e-02]
anha:  70%|######9   | 2309/3300 [11:31<43:41,  2.65s/it, lr: 1.0e-04 loss: 5.998e-02]
Saving at step 2310
Saved checkpoint to /workspace/ai-toolkit/output/anha/anha_000002310.safetensors
Saved optimizer to /workspace/ai-toolkit/output/anha/optimizer.pt

anha:  70%|#######   | 2310/3300 [11:32<43:38,  2.65s/it, lr: 1.0e-04 loss: 6.085e-02]
anha:  70%|#######   | 2310/3300 [11:32<43:38,  2.65s/it, lr: 1.0e-04 loss: 6.085e-02]
anha:  70%|#######   | 2311/3300 [11:32<28:14,  1.71s/it, lr: 1.0e-04 loss: 6.085e-02]
anha:  70%|#######   | 2311/3300 [11:32<28:14,  1.71s/it, lr: 1.0e-04 loss: 6.085e-02]
anha:  70%|#######   | 2311/3300 [11:33<28:14,  1.71s/it, lr: 1.0e-04 loss: 4.119e-02]
anha:  70%|#######   | 2311/3300 [11:33<28:14,  1.71s/it, lr: 1.0e-04 loss: 4.119e-02]
anha:  70%|#######   | 2312/3300 [11:33<25:20,  1.54s/it, lr: 1.0e-04 loss: 4.119e-02]
anha:  70%|#######   | 2312/3300 [11:33<25:20,  1.54s/it, lr: 1.0e-04 loss: 4.119e-02]
anha:  70%|#######   | 2312/3300 [11:37<25:20,  1.54s/it, lr: 1.0e-04 loss: 4.539e-02]
anha:  70%|#######   | 2312/3300 [11:37<25:20,  1.54s/it, lr: 1.0e-04 loss: 4.539e-02]
anha:  70%|#######   | 2313/3300 [11:37<33:38,  2.04s/it, lr: 1.0e-04 loss: 4.539e-02]
anha:  70%|#######   | 2313/3300 [11:37<33:38,  2.04s/it, lr: 1.0e-04 loss: 4.539e-02]
anha:  70%|#######   | 2313/3300 [11:38<33:38,  2.04s/it, lr: 1.0e-04 loss: 1.504e-01]
anha:  70%|#######   | 2313/3300 [11:38<33:38,  2.04s/it, lr: 1.0e-04 loss: 1.504e-01]
anha:  70%|#######   | 2314/3300 [11:38<32:11,  1.96s/it, lr: 1.0e-04 loss: 1.504e-01]
anha:  70%|#######   | 2314/3300 [11:38<32:11,  1.96s/it, lr: 1.0e-04 loss: 1.504e-01]
anha:  70%|#######   | 2314/3300 [11:39<32:11,  1.96s/it, lr: 1.0e-04 loss: 4.986e-02]
anha:  70%|#######   | 2314/3300 [11:39<32:11,  1.96s/it, lr: 1.0e-04 loss: 4.986e-02]
anha:  70%|#######   | 2315/3300 [11:39<27:48,  1.69s/it, lr: 1.0e-04 loss: 4.986e-02]
anha:  70%|#######   | 2315/3300 [11:39<27:48,  1.69s/it, lr: 1.0e-04 loss: 4.986e-02]
anha:  70%|#######   | 2315/3300 [11:41<27:48,  1.69s/it, lr: 1.0e-04 loss: 4.549e-02]
anha:  70%|#######   | 2315/3300 [11:41<27:48,  1.69s/it, lr: 1.0e-04 loss: 4.549e-02]
anha:  70%|#######   | 2316/3300 [11:41<29:06,  1.78s/it, lr: 1.0e-04 loss: 4.549e-02]
anha:  70%|#######   | 2316/3300 [11:41<29:06,  1.78s/it, lr: 1.0e-04 loss: 4.549e-02]
anha:  70%|#######   | 2316/3300 [11:43<29:06,  1.78s/it, lr: 1.0e-04 loss: 5.089e-02]
anha:  70%|#######   | 2316/3300 [11:43<29:06,  1.78s/it, lr: 1.0e-04 loss: 5.089e-02]
anha:  70%|#######   | 2317/3300 [11:43<29:59,  1.83s/it, lr: 1.0e-04 loss: 5.089e-02]
anha:  70%|#######   | 2317/3300 [11:43<29:59,  1.83s/it, lr: 1.0e-04 loss: 5.089e-02]
anha:  70%|#######   | 2317/3300 [11:46<29:59,  1.83s/it, lr: 1.0e-04 loss: 3.161e-02]
anha:  70%|#######   | 2317/3300 [11:46<29:59,  1.83s/it, lr: 1.0e-04 loss: 3.161e-02]
anha:  70%|#######   | 2318/3300 [11:46<36:46,  2.25s/it, lr: 1.0e-04 loss: 3.161e-02]
anha:  70%|#######   | 2318/3300 [11:46<36:46,  2.25s/it, lr: 1.0e-04 loss: 3.161e-02]
anha:  70%|#######   | 2318/3300 [11:48<36:46,  2.25s/it, lr: 1.0e-04 loss: 1.110e-01]
anha:  70%|#######   | 2318/3300 [11:48<36:46,  2.25s/it, lr: 1.0e-04 loss: 1.110e-01]
anha:  70%|#######   | 2319/3300 [11:48<35:26,  2.17s/it, lr: 1.0e-04 loss: 1.110e-01]
anha:  70%|#######   | 2319/3300 [11:48<35:26,  2.17s/it, lr: 1.0e-04 loss: 1.110e-01]
anha:  70%|#######   | 2319/3300 [11:50<35:26,  2.17s/it, lr: 1.0e-04 loss: 4.793e-02]
anha:  70%|#######   | 2319/3300 [11:50<35:26,  2.17s/it, lr: 1.0e-04 loss: 4.793e-02]
anha:  70%|#######   | 2320/3300 [11:51<35:24,  2.17s/it, lr: 1.0e-04 loss: 2.074e-01]
anha:  70%|#######   | 2320/3300 [11:51<35:24,  2.17s/it, lr: 1.0e-04 loss: 2.074e-01]
anha:  70%|#######   | 2321/3300 [11:51<22:56,  1.41s/it, lr: 1.0e-04 loss: 2.074e-01]
anha:  70%|#######   | 2321/3300 [11:51<22:56,  1.41s/it, lr: 1.0e-04 loss: 2.074e-01]
anha:  70%|#######   | 2321/3300 [11:54<22:56,  1.41s/it, lr: 1.0e-04 loss: 8.940e-02]
anha:  70%|#######   | 2321/3300 [11:54<22:56,  1.41s/it, lr: 1.0e-04 loss: 8.940e-02]
anha:  70%|#######   | 2322/3300 [11:54<31:14,  1.92s/it, lr: 1.0e-04 loss: 8.940e-02]
anha:  70%|#######   | 2322/3300 [11:54<31:14,  1.92s/it, lr: 1.0e-04 loss: 8.940e-02]
anha:  70%|#######   | 2322/3300 [11:55<31:14,  1.92s/it, lr: 1.0e-04 loss: 1.206e-01]
anha:  70%|#######   | 2322/3300 [11:55<31:14,  1.92s/it, lr: 1.0e-04 loss: 1.206e-01]
anha:  70%|#######   | 2323/3300 [11:55<27:17,  1.68s/it, lr: 1.0e-04 loss: 1.206e-01]
anha:  70%|#######   | 2323/3300 [11:55<27:17,  1.68s/it, lr: 1.0e-04 loss: 1.206e-01]
anha:  70%|#######   | 2323/3300 [11:57<27:17,  1.68s/it, lr: 1.0e-04 loss: 7.568e-02]
anha:  70%|#######   | 2323/3300 [11:57<27:17,  1.68s/it, lr: 1.0e-04 loss: 7.568e-02]
anha:  70%|#######   | 2324/3300 [11:57<28:35,  1.76s/it, lr: 1.0e-04 loss: 7.568e-02]
anha:  70%|#######   | 2324/3300 [11:57<28:35,  1.76s/it, lr: 1.0e-04 loss: 7.568e-02]
anha:  70%|#######   | 2324/3300 [11:58<28:35,  1.76s/it, lr: 1.0e-04 loss: 9.386e-02]
anha:  70%|#######   | 2324/3300 [11:58<28:35,  1.76s/it, lr: 1.0e-04 loss: 9.386e-02]
anha:  70%|#######   | 2325/3300 [11:58<25:09,  1.55s/it, lr: 1.0e-04 loss: 9.386e-02]
anha:  70%|#######   | 2325/3300 [11:58<25:09,  1.55s/it, lr: 1.0e-04 loss: 9.386e-02]
anha:  70%|#######   | 2325/3300 [12:01<25:09,  1.55s/it, lr: 1.0e-04 loss: 9.403e-02]
anha:  70%|#######   | 2325/3300 [12:01<25:09,  1.55s/it, lr: 1.0e-04 loss: 9.403e-02]
anha:  70%|#######   | 2326/3300 [12:01<33:34,  2.07s/it, lr: 1.0e-04 loss: 9.403e-02]
anha:  70%|#######   | 2326/3300 [12:01<33:34,  2.07s/it, lr: 1.0e-04 loss: 9.403e-02]
anha:  70%|#######   | 2326/3300 [12:02<33:34,  2.07s/it, lr: 1.0e-04 loss: 6.713e-02]
anha:  70%|#######   | 2326/3300 [12:02<33:34,  2.07s/it, lr: 1.0e-04 loss: 6.713e-02]
anha:  71%|#######   | 2327/3300 [12:02<28:37,  1.77s/it, lr: 1.0e-04 loss: 6.713e-02]
anha:  71%|#######   | 2327/3300 [12:02<28:37,  1.77s/it, lr: 1.0e-04 loss: 6.713e-02]
anha:  71%|#######   | 2327/3300 [12:06<28:37,  1.77s/it, lr: 1.0e-04 loss: 7.326e-02]
anha:  71%|#######   | 2327/3300 [12:06<28:37,  1.77s/it, lr: 1.0e-04 loss: 7.326e-02]
anha:  71%|#######   | 2328/3300 [12:06<35:35,  2.20s/it, lr: 1.0e-04 loss: 7.326e-02]
anha:  71%|#######   | 2328/3300 [12:06<35:35,  2.20s/it, lr: 1.0e-04 loss: 7.326e-02]
anha:  71%|#######   | 2328/3300 [12:08<35:35,  2.20s/it, lr: 1.0e-04 loss: 9.254e-02]
anha:  71%|#######   | 2328/3300 [12:08<35:35,  2.20s/it, lr: 1.0e-04 loss: 9.254e-02]
anha:  71%|#######   | 2329/3300 [12:08<34:48,  2.15s/it, lr: 1.0e-04 loss: 9.254e-02]
anha:  71%|#######   | 2329/3300 [12:08<34:48,  2.15s/it, lr: 1.0e-04 loss: 9.254e-02]
anha:  71%|#######   | 2329/3300 [12:10<34:48,  2.15s/it, lr: 1.0e-04 loss: 9.809e-02]
anha:  71%|#######   | 2329/3300 [12:10<34:48,  2.15s/it, lr: 1.0e-04 loss: 9.809e-02]
anha:  71%|#######   | 2330/3300 [12:12<34:45,  2.15s/it, lr: 1.0e-04 loss: 6.002e-02]
anha:  71%|#######   | 2330/3300 [12:12<34:45,  2.15s/it, lr: 1.0e-04 loss: 6.002e-02]
anha:  71%|#######   | 2331/3300 [12:12<26:08,  1.62s/it, lr: 1.0e-04 loss: 6.002e-02]
anha:  71%|#######   | 2331/3300 [12:12<26:08,  1.62s/it, lr: 1.0e-04 loss: 6.002e-02]
anha:  71%|#######   | 2331/3300 [12:15<26:08,  1.62s/it, lr: 1.0e-04 loss: 3.845e-02]
anha:  71%|#######   | 2331/3300 [12:15<26:08,  1.62s/it, lr: 1.0e-04 loss: 3.845e-02]
anha:  71%|#######   | 2332/3300 [12:15<33:30,  2.08s/it, lr: 1.0e-04 loss: 3.845e-02]
anha:  71%|#######   | 2332/3300 [12:15<33:30,  2.08s/it, lr: 1.0e-04 loss: 3.845e-02]
anha:  71%|#######   | 2332/3300 [12:18<33:30,  2.08s/it, lr: 1.0e-04 loss: 5.234e-02]
anha:  71%|#######   | 2332/3300 [12:18<33:30,  2.08s/it, lr: 1.0e-04 loss: 5.234e-02]
anha:  71%|#######   | 2333/3300 [12:18<38:05,  2.36s/it, lr: 1.0e-04 loss: 5.234e-02]
anha:  71%|#######   | 2333/3300 [12:18<38:05,  2.36s/it, lr: 1.0e-04 loss: 5.234e-02]
anha:  71%|#######   | 2333/3300 [12:22<38:05,  2.36s/it, lr: 1.0e-04 loss: 3.357e-02]
anha:  71%|#######   | 2333/3300 [12:22<38:05,  2.36s/it, lr: 1.0e-04 loss: 3.357e-02]
anha:  71%|#######   | 2334/3300 [12:22<42:55,  2.67s/it, lr: 1.0e-04 loss: 3.357e-02]
anha:  71%|#######   | 2334/3300 [12:22<42:55,  2.67s/it, lr: 1.0e-04 loss: 3.357e-02]
anha:  71%|#######   | 2334/3300 [12:23<42:55,  2.67s/it, lr: 1.0e-04 loss: 6.606e-02]
anha:  71%|#######   | 2334/3300 [12:23<42:55,  2.67s/it, lr: 1.0e-04 loss: 6.606e-02]
anha:  71%|#######   | 2335/3300 [12:23<35:24,  2.20s/it, lr: 1.0e-04 loss: 6.606e-02]
anha:  71%|#######   | 2335/3300 [12:23<35:24,  2.20s/it, lr: 1.0e-04 loss: 6.606e-02]
anha:  71%|#######   | 2335/3300 [12:25<35:24,  2.20s/it, lr: 1.0e-04 loss: 6.150e-02]
anha:  71%|#######   | 2335/3300 [12:25<35:24,  2.20s/it, lr: 1.0e-04 loss: 6.150e-02]
anha:  71%|#######   | 2336/3300 [12:25<34:20,  2.14s/it, lr: 1.0e-04 loss: 6.150e-02]
anha:  71%|#######   | 2336/3300 [12:25<34:20,  2.14s/it, lr: 1.0e-04 loss: 6.150e-02]
anha:  71%|#######   | 2336/3300 [12:28<34:20,  2.14s/it, lr: 1.0e-04 loss: 6.562e-02]
anha:  71%|#######   | 2336/3300 [12:28<34:20,  2.14s/it, lr: 1.0e-04 loss: 6.562e-02]
anha:  71%|#######   | 2337/3300 [12:28<39:38,  2.47s/it, lr: 1.0e-04 loss: 6.562e-02]
anha:  71%|#######   | 2337/3300 [12:28<39:38,  2.47s/it, lr: 1.0e-04 loss: 6.562e-02]
anha:  71%|#######   | 2337/3300 [12:30<39:38,  2.47s/it, lr: 1.0e-04 loss: 3.377e-02]
anha:  71%|#######   | 2337/3300 [12:30<39:38,  2.47s/it, lr: 1.0e-04 loss: 3.377e-02]
anha:  71%|#######   | 2338/3300 [12:30<36:09,  2.26s/it, lr: 1.0e-04 loss: 3.377e-02]
anha:  71%|#######   | 2338/3300 [12:30<36:09,  2.26s/it, lr: 1.0e-04 loss: 3.377e-02]
anha:  71%|#######   | 2338/3300 [12:32<36:09,  2.26s/it, lr: 1.0e-04 loss: 5.137e-02]
anha:  71%|#######   | 2338/3300 [12:32<36:09,  2.26s/it, lr: 1.0e-04 loss: 5.137e-02]
anha:  71%|#######   | 2339/3300 [12:32<35:05,  2.19s/it, lr: 1.0e-04 loss: 5.137e-02]
anha:  71%|#######   | 2339/3300 [12:32<35:05,  2.19s/it, lr: 1.0e-04 loss: 5.137e-02]
anha:  71%|#######   | 2339/3300 [12:35<35:05,  2.19s/it, lr: 1.0e-04 loss: 3.924e-01]
anha:  71%|#######   | 2339/3300 [12:35<35:05,  2.19s/it, lr: 1.0e-04 loss: 3.924e-01]
anha:  71%|#######   | 2340/3300 [12:37<35:02,  2.19s/it, lr: 1.0e-04 loss: 9.391e-02]
anha:  71%|#######   | 2340/3300 [12:37<35:02,  2.19s/it, lr: 1.0e-04 loss: 9.391e-02]
anha:  71%|#######   | 2341/3300 [12:37<25:16,  1.58s/it, lr: 1.0e-04 loss: 9.391e-02]
anha:  71%|#######   | 2341/3300 [12:37<25:16,  1.58s/it, lr: 1.0e-04 loss: 9.391e-02]
anha:  71%|#######   | 2341/3300 [12:39<25:16,  1.58s/it, lr: 1.0e-04 loss: 4.143e-02]
anha:  71%|#######   | 2341/3300 [12:39<25:16,  1.58s/it, lr: 1.0e-04 loss: 4.143e-02]
anha:  71%|#######   | 2342/3300 [12:39<25:52,  1.62s/it, lr: 1.0e-04 loss: 4.143e-02]
anha:  71%|#######   | 2342/3300 [12:39<25:52,  1.62s/it, lr: 1.0e-04 loss: 4.143e-02]
anha:  71%|#######   | 2342/3300 [12:42<25:52,  1.62s/it, lr: 1.0e-04 loss: 5.769e-02]
anha:  71%|#######   | 2342/3300 [12:42<25:52,  1.62s/it, lr: 1.0e-04 loss: 5.769e-02]
anha:  71%|#######1  | 2343/3300 [12:42<33:05,  2.08s/it, lr: 1.0e-04 loss: 5.769e-02]
anha:  71%|#######1  | 2343/3300 [12:42<33:05,  2.08s/it, lr: 1.0e-04 loss: 5.769e-02]
anha:  71%|#######1  | 2343/3300 [12:44<33:05,  2.08s/it, lr: 1.0e-04 loss: 9.240e-02]
anha:  71%|#######1  | 2343/3300 [12:44<33:05,  2.08s/it, lr: 1.0e-04 loss: 9.240e-02]
anha:  71%|#######1  | 2344/3300 [12:44<32:38,  2.05s/it, lr: 1.0e-04 loss: 9.240e-02]
anha:  71%|#######1  | 2344/3300 [12:44<32:38,  2.05s/it, lr: 1.0e-04 loss: 9.240e-02]
anha:  71%|#######1  | 2344/3300 [12:47<32:38,  2.05s/it, lr: 1.0e-04 loss: 2.120e-01]
anha:  71%|#######1  | 2344/3300 [12:47<32:38,  2.05s/it, lr: 1.0e-04 loss: 2.120e-01]
anha:  71%|#######1  | 2345/3300 [12:47<38:06,  2.39s/it, lr: 1.0e-04 loss: 2.120e-01]
anha:  71%|#######1  | 2345/3300 [12:47<38:06,  2.39s/it, lr: 1.0e-04 loss: 2.120e-01]
anha:  71%|#######1  | 2345/3300 [12:48<38:06,  2.39s/it, lr: 1.0e-04 loss: 4.685e-01]
anha:  71%|#######1  | 2345/3300 [12:48<38:06,  2.39s/it, lr: 1.0e-04 loss: 4.685e-01]
anha:  71%|#######1  | 2346/3300 [12:48<31:43,  2.00s/it, lr: 1.0e-04 loss: 4.685e-01]
anha:  71%|#######1  | 2346/3300 [12:48<31:43,  2.00s/it, lr: 1.0e-04 loss: 4.685e-01]
anha:  71%|#######1  | 2346/3300 [12:49<31:43,  2.00s/it, lr: 1.0e-04 loss: 8.725e-02]
anha:  71%|#######1  | 2346/3300 [12:49<31:43,  2.00s/it, lr: 1.0e-04 loss: 8.725e-02]
anha:  71%|#######1  | 2347/3300 [12:49<27:08,  1.71s/it, lr: 1.0e-04 loss: 8.725e-02]
anha:  71%|#######1  | 2347/3300 [12:49<27:08,  1.71s/it, lr: 1.0e-04 loss: 8.725e-02]
anha:  71%|#######1  | 2347/3300 [12:50<27:08,  1.71s/it, lr: 1.0e-04 loss: 3.188e-02]
anha:  71%|#######1  | 2347/3300 [12:50<27:08,  1.71s/it, lr: 1.0e-04 loss: 3.188e-02]
anha:  71%|#######1  | 2348/3300 [12:50<24:20,  1.53s/it, lr: 1.0e-04 loss: 3.188e-02]
anha:  71%|#######1  | 2348/3300 [12:50<24:20,  1.53s/it, lr: 1.0e-04 loss: 3.188e-02]
anha:  71%|#######1  | 2348/3300 [12:52<24:20,  1.53s/it, lr: 1.0e-04 loss: 2.867e-02]
anha:  71%|#######1  | 2348/3300 [12:52<24:20,  1.53s/it, lr: 1.0e-04 loss: 2.867e-02]
anha:  71%|#######1  | 2349/3300 [12:52<26:16,  1.66s/it, lr: 1.0e-04 loss: 2.867e-02]
anha:  71%|#######1  | 2349/3300 [12:52<26:16,  1.66s/it, lr: 1.0e-04 loss: 2.867e-02]
anha:  71%|#######1  | 2349/3300 [12:54<26:16,  1.66s/it, lr: 1.0e-04 loss: 1.248e-01]
anha:  71%|#######1  | 2349/3300 [12:54<26:16,  1.66s/it, lr: 1.0e-04 loss: 1.248e-01]
anha:  71%|#######1  | 2350/3300 [12:55<26:14,  1.66s/it, lr: 1.0e-04 loss: 5.133e-02]
anha:  71%|#######1  | 2350/3300 [12:55<26:14,  1.66s/it, lr: 1.0e-04 loss: 5.133e-02]
anha:  71%|#######1  | 2351/3300 [12:55<17:53,  1.13s/it, lr: 1.0e-04 loss: 5.133e-02]
anha:  71%|#######1  | 2351/3300 [12:55<17:53,  1.13s/it, lr: 1.0e-04 loss: 5.133e-02]
anha:  71%|#######1  | 2351/3300 [12:58<17:53,  1.13s/it, lr: 1.0e-04 loss: 1.423e-01]
anha:  71%|#######1  | 2351/3300 [12:58<17:53,  1.13s/it, lr: 1.0e-04 loss: 1.423e-01]
anha:  71%|#######1  | 2352/3300 [12:58<26:34,  1.68s/it, lr: 1.0e-04 loss: 1.423e-01]
anha:  71%|#######1  | 2352/3300 [12:58<26:34,  1.68s/it, lr: 1.0e-04 loss: 1.423e-01]
anha:  71%|#######1  | 2352/3300 [13:00<26:34,  1.68s/it, lr: 1.0e-04 loss: 4.899e-02]
anha:  71%|#######1  | 2352/3300 [13:00<26:34,  1.68s/it, lr: 1.0e-04 loss: 4.899e-02]
anha:  71%|#######1  | 2353/3300 [13:00<27:39,  1.75s/it, lr: 1.0e-04 loss: 4.899e-02]
anha:  71%|#######1  | 2353/3300 [13:00<27:39,  1.75s/it, lr: 1.0e-04 loss: 4.899e-02]
anha:  71%|#######1  | 2353/3300 [13:02<27:39,  1.75s/it, lr: 1.0e-04 loss: 8.871e-02]
anha:  71%|#######1  | 2353/3300 [13:02<27:39,  1.75s/it, lr: 1.0e-04 loss: 8.871e-02]
anha:  71%|#######1  | 2354/3300 [13:02<27:34,  1.75s/it, lr: 1.0e-04 loss: 8.871e-02]
anha:  71%|#######1  | 2354/3300 [13:02<27:34,  1.75s/it, lr: 1.0e-04 loss: 8.871e-02]
anha:  71%|#######1  | 2354/3300 [13:03<27:34,  1.75s/it, lr: 1.0e-04 loss: 3.894e-02]
anha:  71%|#######1  | 2354/3300 [13:03<27:34,  1.75s/it, lr: 1.0e-04 loss: 3.894e-02]
anha:  71%|#######1  | 2355/3300 [13:03<24:28,  1.55s/it, lr: 1.0e-04 loss: 3.894e-02]
anha:  71%|#######1  | 2355/3300 [13:03<24:28,  1.55s/it, lr: 1.0e-04 loss: 3.894e-02]
anha:  71%|#######1  | 2355/3300 [13:04<24:28,  1.55s/it, lr: 1.0e-04 loss: 6.180e-02]
anha:  71%|#######1  | 2355/3300 [13:04<24:28,  1.55s/it, lr: 1.0e-04 loss: 6.180e-02]
anha:  71%|#######1  | 2356/3300 [13:04<22:00,  1.40s/it, lr: 1.0e-04 loss: 6.180e-02]
anha:  71%|#######1  | 2356/3300 [13:04<22:00,  1.40s/it, lr: 1.0e-04 loss: 6.180e-02]
anha:  71%|#######1  | 2356/3300 [13:08<22:00,  1.40s/it, lr: 1.0e-04 loss: 2.761e-02]
anha:  71%|#######1  | 2356/3300 [13:08<22:00,  1.40s/it, lr: 1.0e-04 loss: 2.761e-02]
anha:  71%|#######1  | 2357/3300 [13:08<31:05,  1.98s/it, lr: 1.0e-04 loss: 2.761e-02]
anha:  71%|#######1  | 2357/3300 [13:08<31:05,  1.98s/it, lr: 1.0e-04 loss: 2.761e-02]
anha:  71%|#######1  | 2357/3300 [13:09<31:05,  1.98s/it, lr: 1.0e-04 loss: 4.003e-01]
anha:  71%|#######1  | 2357/3300 [13:09<31:05,  1.98s/it, lr: 1.0e-04 loss: 4.003e-01]
anha:  71%|#######1  | 2358/3300 [13:09<30:56,  1.97s/it, lr: 1.0e-04 loss: 4.003e-01]
anha:  71%|#######1  | 2358/3300 [13:09<30:56,  1.97s/it, lr: 1.0e-04 loss: 4.003e-01]
anha:  71%|#######1  | 2358/3300 [13:10<30:56,  1.97s/it, lr: 1.0e-04 loss: 1.100e-01]
anha:  71%|#######1  | 2358/3300 [13:10<30:56,  1.97s/it, lr: 1.0e-04 loss: 1.100e-01]
anha:  71%|#######1  | 2359/3300 [13:10<26:26,  1.69s/it, lr: 1.0e-04 loss: 1.100e-01]
anha:  71%|#######1  | 2359/3300 [13:10<26:26,  1.69s/it, lr: 1.0e-04 loss: 1.100e-01]
anha:  71%|#######1  | 2359/3300 [13:14<26:26,  1.69s/it, lr: 1.0e-04 loss: 5.489e-02]
anha:  71%|#######1  | 2359/3300 [13:14<26:26,  1.69s/it, lr: 1.0e-04 loss: 5.489e-02]
anha:  72%|#######1  | 2360/3300 [13:17<26:24,  1.69s/it, lr: 1.0e-04 loss: 7.512e-02]
anha:  72%|#######1  | 2360/3300 [13:17<26:24,  1.69s/it, lr: 1.0e-04 loss: 7.512e-02]
anha:  72%|#######1  | 2361/3300 [13:17<26:05,  1.67s/it, lr: 1.0e-04 loss: 7.512e-02]
anha:  72%|#######1  | 2361/3300 [13:17<26:05,  1.67s/it, lr: 1.0e-04 loss: 7.512e-02]
anha:  72%|#######1  | 2361/3300 [13:18<26:05,  1.67s/it, lr: 1.0e-04 loss: 1.096e-01]
anha:  72%|#######1  | 2361/3300 [13:18<26:05,  1.67s/it, lr: 1.0e-04 loss: 1.096e-01]
anha:  72%|#######1  | 2362/3300 [13:18<23:31,  1.51s/it, lr: 1.0e-04 loss: 1.096e-01]
anha:  72%|#######1  | 2362/3300 [13:18<23:31,  1.51s/it, lr: 1.0e-04 loss: 1.096e-01]
anha:  72%|#######1  | 2362/3300 [13:20<23:31,  1.51s/it, lr: 1.0e-04 loss: 4.518e-02]
anha:  72%|#######1  | 2362/3300 [13:20<23:31,  1.51s/it, lr: 1.0e-04 loss: 4.518e-02]
anha:  72%|#######1  | 2363/3300 [13:20<25:19,  1.62s/it, lr: 1.0e-04 loss: 4.518e-02]
anha:  72%|#######1  | 2363/3300 [13:20<25:19,  1.62s/it, lr: 1.0e-04 loss: 4.518e-02]
anha:  72%|#######1  | 2363/3300 [13:24<25:19,  1.62s/it, lr: 1.0e-04 loss: 1.046e-01]
anha:  72%|#######1  | 2363/3300 [13:24<25:19,  1.62s/it, lr: 1.0e-04 loss: 1.046e-01]
anha:  72%|#######1  | 2364/3300 [13:24<33:08,  2.12s/it, lr: 1.0e-04 loss: 1.046e-01]
anha:  72%|#######1  | 2364/3300 [13:24<33:08,  2.12s/it, lr: 1.0e-04 loss: 1.046e-01]
anha:  72%|#######1  | 2364/3300 [13:27<33:08,  2.12s/it, lr: 1.0e-04 loss: 4.443e-02]
anha:  72%|#######1  | 2364/3300 [13:27<33:08,  2.12s/it, lr: 1.0e-04 loss: 4.443e-02]
anha:  72%|#######1  | 2365/3300 [13:27<38:50,  2.49s/it, lr: 1.0e-04 loss: 4.443e-02]
anha:  72%|#######1  | 2365/3300 [13:27<38:50,  2.49s/it, lr: 1.0e-04 loss: 4.443e-02]
anha:  72%|#######1  | 2365/3300 [13:29<38:50,  2.49s/it, lr: 1.0e-04 loss: 6.813e-02]
anha:  72%|#######1  | 2365/3300 [13:29<38:50,  2.49s/it, lr: 1.0e-04 loss: 6.813e-02]
anha:  72%|#######1  | 2366/3300 [13:29<36:23,  2.34s/it, lr: 1.0e-04 loss: 6.813e-02]
anha:  72%|#######1  | 2366/3300 [13:29<36:23,  2.34s/it, lr: 1.0e-04 loss: 6.813e-02]
anha:  72%|#######1  | 2366/3300 [13:30<36:23,  2.34s/it, lr: 1.0e-04 loss: 1.708e-01]
anha:  72%|#######1  | 2366/3300 [13:30<36:23,  2.34s/it, lr: 1.0e-04 loss: 1.708e-01]
anha:  72%|#######1  | 2367/3300 [13:30<30:53,  1.99s/it, lr: 1.0e-04 loss: 1.708e-01]
anha:  72%|#######1  | 2367/3300 [13:30<30:53,  1.99s/it, lr: 1.0e-04 loss: 1.708e-01]
anha:  72%|#######1  | 2367/3300 [13:34<30:53,  1.99s/it, lr: 1.0e-04 loss: 7.632e-02]
anha:  72%|#######1  | 2367/3300 [13:34<30:53,  1.99s/it, lr: 1.0e-04 loss: 7.632e-02]
anha:  72%|#######1  | 2368/3300 [13:34<37:38,  2.42s/it, lr: 1.0e-04 loss: 7.632e-02]
anha:  72%|#######1  | 2368/3300 [13:34<37:38,  2.42s/it, lr: 1.0e-04 loss: 7.632e-02]
anha:  72%|#######1  | 2368/3300 [13:37<37:38,  2.42s/it, lr: 1.0e-04 loss: 3.555e-02]
anha:  72%|#######1  | 2368/3300 [13:37<37:38,  2.42s/it, lr: 1.0e-04 loss: 3.555e-02]
anha:  72%|#######1  | 2369/3300 [13:37<41:57,  2.70s/it, lr: 1.0e-04 loss: 3.555e-02]
anha:  72%|#######1  | 2369/3300 [13:37<41:57,  2.70s/it, lr: 1.0e-04 loss: 3.555e-02]
anha:  72%|#######1  | 2369/3300 [13:40<41:57,  2.70s/it, lr: 1.0e-04 loss: 9.298e-02]
anha:  72%|#######1  | 2369/3300 [13:40<41:57,  2.70s/it, lr: 1.0e-04 loss: 9.298e-02]
anha:  72%|#######1  | 2370/3300 [13:41<41:54,  2.70s/it, lr: 1.0e-04 loss: 5.622e-02]
anha:  72%|#######1  | 2370/3300 [13:41<41:54,  2.70s/it, lr: 1.0e-04 loss: 5.622e-02]
anha:  72%|#######1  | 2371/3300 [13:41<26:16,  1.70s/it, lr: 1.0e-04 loss: 5.622e-02]
anha:  72%|#######1  | 2371/3300 [13:41<26:16,  1.70s/it, lr: 1.0e-04 loss: 5.622e-02]
anha:  72%|#######1  | 2371/3300 [13:42<26:16,  1.70s/it, lr: 1.0e-04 loss: 7.289e-02]
anha:  72%|#######1  | 2371/3300 [13:42<26:16,  1.70s/it, lr: 1.0e-04 loss: 7.289e-02]
anha:  72%|#######1  | 2372/3300 [13:42<23:34,  1.52s/it, lr: 1.0e-04 loss: 7.289e-02]
anha:  72%|#######1  | 2372/3300 [13:42<23:34,  1.52s/it, lr: 1.0e-04 loss: 7.289e-02]
anha:  72%|#######1  | 2372/3300 [13:43<23:34,  1.52s/it, lr: 1.0e-04 loss: 7.094e-02]
anha:  72%|#######1  | 2372/3300 [13:43<23:34,  1.52s/it, lr: 1.0e-04 loss: 7.094e-02]
anha:  72%|#######1  | 2373/3300 [13:43<21:25,  1.39s/it, lr: 1.0e-04 loss: 7.094e-02]
anha:  72%|#######1  | 2373/3300 [13:43<21:25,  1.39s/it, lr: 1.0e-04 loss: 7.094e-02]
anha:  72%|#######1  | 2373/3300 [13:46<21:25,  1.39s/it, lr: 1.0e-04 loss: 3.225e-02]
anha:  72%|#######1  | 2373/3300 [13:46<21:25,  1.39s/it, lr: 1.0e-04 loss: 3.225e-02]
anha:  72%|#######1  | 2374/3300 [13:46<26:05,  1.69s/it, lr: 1.0e-04 loss: 3.225e-02]
anha:  72%|#######1  | 2374/3300 [13:46<26:05,  1.69s/it, lr: 1.0e-04 loss: 3.225e-02]
anha:  72%|#######1  | 2374/3300 [13:47<26:05,  1.69s/it, lr: 1.0e-04 loss: 6.464e-02]
anha:  72%|#######1  | 2374/3300 [13:47<26:05,  1.69s/it, lr: 1.0e-04 loss: 6.464e-02]
anha:  72%|#######1  | 2375/3300 [13:47<23:06,  1.50s/it, lr: 1.0e-04 loss: 6.464e-02]
anha:  72%|#######1  | 2375/3300 [13:47<23:06,  1.50s/it, lr: 1.0e-04 loss: 6.464e-02]
anha:  72%|#######1  | 2375/3300 [13:48<23:06,  1.50s/it, lr: 1.0e-04 loss: 4.676e-02]
anha:  72%|#######1  | 2375/3300 [13:48<23:06,  1.50s/it, lr: 1.0e-04 loss: 4.676e-02]
anha:  72%|#######2  | 2376/3300 [13:48<21:37,  1.40s/it, lr: 1.0e-04 loss: 4.676e-02]
anha:  72%|#######2  | 2376/3300 [13:48<21:37,  1.40s/it, lr: 1.0e-04 loss: 4.676e-02]
anha:  72%|#######2  | 2376/3300 [13:49<21:37,  1.40s/it, lr: 1.0e-04 loss: 7.203e-02]
anha:  72%|#######2  | 2376/3300 [13:49<21:37,  1.40s/it, lr: 1.0e-04 loss: 7.203e-02]
anha:  72%|#######2  | 2377/3300 [13:49<19:48,  1.29s/it, lr: 1.0e-04 loss: 7.203e-02]
anha:  72%|#######2  | 2377/3300 [13:49<19:48,  1.29s/it, lr: 1.0e-04 loss: 7.203e-02]
anha:  72%|#######2  | 2377/3300 [13:50<19:48,  1.29s/it, lr: 1.0e-04 loss: 7.295e-02]
anha:  72%|#######2  | 2377/3300 [13:50<19:48,  1.29s/it, lr: 1.0e-04 loss: 7.295e-02]
anha:  72%|#######2  | 2378/3300 [13:50<18:28,  1.20s/it, lr: 1.0e-04 loss: 7.295e-02]
anha:  72%|#######2  | 2378/3300 [13:50<18:28,  1.20s/it, lr: 1.0e-04 loss: 7.295e-02]
anha:  72%|#######2  | 2378/3300 [13:51<18:28,  1.20s/it, lr: 1.0e-04 loss: 6.225e-02]
anha:  72%|#######2  | 2378/3300 [13:51<18:28,  1.20s/it, lr: 1.0e-04 loss: 6.225e-02]
anha:  72%|#######2  | 2379/3300 [13:51<17:34,  1.15s/it, lr: 1.0e-04 loss: 6.225e-02]
anha:  72%|#######2  | 2379/3300 [13:51<17:34,  1.15s/it, lr: 1.0e-04 loss: 6.225e-02]
anha:  72%|#######2  | 2379/3300 [13:55<17:34,  1.15s/it, lr: 1.0e-04 loss: 7.739e-02]
anha:  72%|#######2  | 2379/3300 [13:55<17:34,  1.15s/it, lr: 1.0e-04 loss: 7.739e-02]
anha:  72%|#######2  | 2380/3300 [13:56<17:33,  1.15s/it, lr: 1.0e-04 loss: 1.683e-01]
anha:  72%|#######2  | 2380/3300 [13:56<17:33,  1.15s/it, lr: 1.0e-04 loss: 1.683e-01]
anha:  72%|#######2  | 2381/3300 [13:56<13:02,  1.17it/s, lr: 1.0e-04 loss: 1.683e-01]
anha:  72%|#######2  | 2381/3300 [13:56<13:02,  1.17it/s, lr: 1.0e-04 loss: 1.683e-01]
anha:  72%|#######2  | 2381/3300 [13:59<13:02,  1.17it/s, lr: 1.0e-04 loss: 4.603e-02]
anha:  72%|#######2  | 2381/3300 [13:59<13:02,  1.17it/s, lr: 1.0e-04 loss: 4.603e-02]
anha:  72%|#######2  | 2382/3300 [13:59<22:56,  1.50s/it, lr: 1.0e-04 loss: 4.603e-02]
anha:  72%|#######2  | 2382/3300 [13:59<22:56,  1.50s/it, lr: 1.0e-04 loss: 4.603e-02]
anha:  72%|#######2  | 2382/3300 [14:01<22:56,  1.50s/it, lr: 1.0e-04 loss: 5.639e-02]
anha:  72%|#######2  | 2382/3300 [14:01<22:56,  1.50s/it, lr: 1.0e-04 loss: 5.639e-02]
anha:  72%|#######2  | 2383/3300 [14:01<23:50,  1.56s/it, lr: 1.0e-04 loss: 5.639e-02]
anha:  72%|#######2  | 2383/3300 [14:01<23:50,  1.56s/it, lr: 1.0e-04 loss: 5.639e-02]
anha:  72%|#######2  | 2383/3300 [14:02<23:50,  1.56s/it, lr: 1.0e-04 loss: 2.298e-01]
anha:  72%|#######2  | 2383/3300 [14:02<23:50,  1.56s/it, lr: 1.0e-04 loss: 2.298e-01]
anha:  72%|#######2  | 2384/3300 [14:02<21:31,  1.41s/it, lr: 1.0e-04 loss: 2.298e-01]
anha:  72%|#######2  | 2384/3300 [14:02<21:31,  1.41s/it, lr: 1.0e-04 loss: 2.298e-01]
anha:  72%|#######2  | 2384/3300 [14:05<21:31,  1.41s/it, lr: 1.0e-04 loss: 4.372e-02]
anha:  72%|#######2  | 2384/3300 [14:05<21:31,  1.41s/it, lr: 1.0e-04 loss: 4.372e-02]
anha:  72%|#######2  | 2385/3300 [14:05<30:01,  1.97s/it, lr: 1.0e-04 loss: 4.372e-02]
anha:  72%|#######2  | 2385/3300 [14:05<30:01,  1.97s/it, lr: 1.0e-04 loss: 4.372e-02]
anha:  72%|#######2  | 2385/3300 [14:09<30:01,  1.97s/it, lr: 1.0e-04 loss: 1.197e-01]
anha:  72%|#######2  | 2385/3300 [14:09<30:01,  1.97s/it, lr: 1.0e-04 loss: 1.197e-01]
anha:  72%|#######2  | 2386/3300 [14:09<36:42,  2.41s/it, lr: 1.0e-04 loss: 1.197e-01]
anha:  72%|#######2  | 2386/3300 [14:09<36:42,  2.41s/it, lr: 1.0e-04 loss: 1.197e-01]
anha:  72%|#######2  | 2386/3300 [14:10<36:42,  2.41s/it, lr: 1.0e-04 loss: 7.880e-02]
anha:  72%|#######2  | 2386/3300 [14:10<36:42,  2.41s/it, lr: 1.0e-04 loss: 7.880e-02]
anha:  72%|#######2  | 2387/3300 [14:10<30:28,  2.00s/it, lr: 1.0e-04 loss: 7.880e-02]
anha:  72%|#######2  | 2387/3300 [14:10<30:28,  2.00s/it, lr: 1.0e-04 loss: 7.880e-02]
anha:  72%|#######2  | 2387/3300 [14:11<30:28,  2.00s/it, lr: 1.0e-04 loss: 7.999e-02]
anha:  72%|#######2  | 2387/3300 [14:11<30:28,  2.00s/it, lr: 1.0e-04 loss: 7.999e-02]
anha:  72%|#######2  | 2388/3300 [14:11<26:00,  1.71s/it, lr: 1.0e-04 loss: 7.999e-02]
anha:  72%|#######2  | 2388/3300 [14:11<26:00,  1.71s/it, lr: 1.0e-04 loss: 7.999e-02]
anha:  72%|#######2  | 2388/3300 [14:13<26:00,  1.71s/it, lr: 1.0e-04 loss: 5.880e-02]
anha:  72%|#######2  | 2388/3300 [14:13<26:00,  1.71s/it, lr: 1.0e-04 loss: 5.880e-02]
anha:  72%|#######2  | 2389/3300 [14:13<27:10,  1.79s/it, lr: 1.0e-04 loss: 5.880e-02]
anha:  72%|#######2  | 2389/3300 [14:13<27:10,  1.79s/it, lr: 1.0e-04 loss: 5.880e-02]
anha:  72%|#######2  | 2389/3300 [14:16<27:10,  1.79s/it, lr: 1.0e-04 loss: 7.690e-02]
anha:  72%|#######2  | 2389/3300 [14:16<27:10,  1.79s/it, lr: 1.0e-04 loss: 7.690e-02]
anha:  72%|#######2  | 2390/3300 [14:18<27:08,  1.79s/it, lr: 1.0e-04 loss: 1.138e-01]
anha:  72%|#######2  | 2390/3300 [14:18<27:08,  1.79s/it, lr: 1.0e-04 loss: 1.138e-01]
anha:  72%|#######2  | 2391/3300 [14:18<21:33,  1.42s/it, lr: 1.0e-04 loss: 1.138e-01]
anha:  72%|#######2  | 2391/3300 [14:18<21:33,  1.42s/it, lr: 1.0e-04 loss: 1.138e-01]
anha:  72%|#######2  | 2391/3300 [14:22<21:33,  1.42s/it, lr: 1.0e-04 loss: 3.720e-02]
anha:  72%|#######2  | 2391/3300 [14:22<21:33,  1.42s/it, lr: 1.0e-04 loss: 3.720e-02]
anha:  72%|#######2  | 2392/3300 [14:22<29:16,  1.93s/it, lr: 1.0e-04 loss: 3.720e-02]
anha:  72%|#######2  | 2392/3300 [14:22<29:16,  1.93s/it, lr: 1.0e-04 loss: 3.720e-02]
anha:  72%|#######2  | 2392/3300 [14:23<29:16,  1.93s/it, lr: 1.0e-04 loss: 2.893e-02]
anha:  72%|#######2  | 2392/3300 [14:23<29:16,  1.93s/it, lr: 1.0e-04 loss: 2.893e-02]
anha:  73%|#######2  | 2393/3300 [14:23<29:23,  1.94s/it, lr: 1.0e-04 loss: 2.893e-02]
anha:  73%|#######2  | 2393/3300 [14:23<29:23,  1.94s/it, lr: 1.0e-04 loss: 2.893e-02]
anha:  73%|#######2  | 2393/3300 [14:25<29:23,  1.94s/it, lr: 1.0e-04 loss: 7.167e-02]
anha:  73%|#######2  | 2393/3300 [14:25<29:23,  1.94s/it, lr: 1.0e-04 loss: 7.167e-02]
anha:  73%|#######2  | 2394/3300 [14:25<28:29,  1.89s/it, lr: 1.0e-04 loss: 7.167e-02]
anha:  73%|#######2  | 2394/3300 [14:25<28:29,  1.89s/it, lr: 1.0e-04 loss: 7.167e-02]
anha:  73%|#######2  | 2394/3300 [14:26<28:29,  1.89s/it, lr: 1.0e-04 loss: 6.514e-02]
anha:  73%|#######2  | 2394/3300 [14:26<28:29,  1.89s/it, lr: 1.0e-04 loss: 6.514e-02]
anha:  73%|#######2  | 2395/3300 [14:26<25:12,  1.67s/it, lr: 1.0e-04 loss: 6.514e-02]
anha:  73%|#######2  | 2395/3300 [14:26<25:12,  1.67s/it, lr: 1.0e-04 loss: 6.514e-02]
anha:  73%|#######2  | 2395/3300 [14:27<25:12,  1.67s/it, lr: 1.0e-04 loss: 4.405e-02]
anha:  73%|#######2  | 2395/3300 [14:27<25:12,  1.67s/it, lr: 1.0e-04 loss: 4.405e-02]
anha:  73%|#######2  | 2396/3300 [14:27<22:18,  1.48s/it, lr: 1.0e-04 loss: 4.405e-02]
anha:  73%|#######2  | 2396/3300 [14:27<22:18,  1.48s/it, lr: 1.0e-04 loss: 4.405e-02]
anha:  73%|#######2  | 2396/3300 [14:29<22:18,  1.48s/it, lr: 1.0e-04 loss: 5.881e-02]
anha:  73%|#######2  | 2396/3300 [14:29<22:18,  1.48s/it, lr: 1.0e-04 loss: 5.881e-02]
anha:  73%|#######2  | 2397/3300 [14:29<24:20,  1.62s/it, lr: 1.0e-04 loss: 5.881e-02]
anha:  73%|#######2  | 2397/3300 [14:29<24:20,  1.62s/it, lr: 1.0e-04 loss: 5.881e-02]
anha:  73%|#######2  | 2397/3300 [14:31<24:20,  1.62s/it, lr: 1.0e-04 loss: 4.535e-02]
anha:  73%|#######2  | 2397/3300 [14:31<24:20,  1.62s/it, lr: 1.0e-04 loss: 4.535e-02]
anha:  73%|#######2  | 2398/3300 [14:31<25:47,  1.72s/it, lr: 1.0e-04 loss: 4.535e-02]
anha:  73%|#######2  | 2398/3300 [14:31<25:47,  1.72s/it, lr: 1.0e-04 loss: 4.535e-02]
anha:  73%|#######2  | 2398/3300 [14:35<25:47,  1.72s/it, lr: 1.0e-04 loss: 6.022e-02]
anha:  73%|#######2  | 2398/3300 [14:35<25:47,  1.72s/it, lr: 1.0e-04 loss: 6.022e-02]
anha:  73%|#######2  | 2399/3300 [14:35<33:05,  2.20s/it, lr: 1.0e-04 loss: 6.022e-02]
anha:  73%|#######2  | 2399/3300 [14:35<33:05,  2.20s/it, lr: 1.0e-04 loss: 6.022e-02]
anha:  73%|#######2  | 2399/3300 [14:36<33:05,  2.20s/it, lr: 1.0e-04 loss: 1.052e-01]
anha:  73%|#######2  | 2399/3300 [14:36<33:05,  2.20s/it, lr: 1.0e-04 loss: 1.052e-01]
anha:  73%|#######2  | 2400/3300 [14:37<33:03,  2.20s/it, lr: 1.0e-04 loss: 8.076e-02]
anha:  73%|#######2  | 2400/3300 [14:37<33:03,  2.20s/it, lr: 1.0e-04 loss: 8.076e-02]
anha:  73%|#######2  | 2401/3300 [14:37<21:18,  1.42s/it, lr: 1.0e-04 loss: 8.076e-02]
anha:  73%|#######2  | 2401/3300 [14:37<21:18,  1.42s/it, lr: 1.0e-04 loss: 8.076e-02]
anha:  73%|#######2  | 2401/3300 [14:38<21:18,  1.42s/it, lr: 1.0e-04 loss: 8.250e-02]
anha:  73%|#######2  | 2401/3300 [14:38<21:18,  1.42s/it, lr: 1.0e-04 loss: 8.250e-02]
anha:  73%|#######2  | 2402/3300 [14:38<19:42,  1.32s/it, lr: 1.0e-04 loss: 8.250e-02]
anha:  73%|#######2  | 2402/3300 [14:38<19:42,  1.32s/it, lr: 1.0e-04 loss: 8.250e-02]
anha:  73%|#######2  | 2402/3300 [14:41<19:42,  1.32s/it, lr: 1.0e-04 loss: 4.667e-02]
anha:  73%|#######2  | 2402/3300 [14:41<19:42,  1.32s/it, lr: 1.0e-04 loss: 4.667e-02]
anha:  73%|#######2  | 2403/3300 [14:41<28:06,  1.88s/it, lr: 1.0e-04 loss: 4.667e-02]
anha:  73%|#######2  | 2403/3300 [14:41<28:06,  1.88s/it, lr: 1.0e-04 loss: 4.667e-02]
anha:  73%|#######2  | 2403/3300 [14:44<28:06,  1.88s/it, lr: 1.0e-04 loss: 2.513e-02]
anha:  73%|#######2  | 2403/3300 [14:44<28:06,  1.88s/it, lr: 1.0e-04 loss: 2.513e-02]
anha:  73%|#######2  | 2404/3300 [14:44<34:18,  2.30s/it, lr: 1.0e-04 loss: 2.513e-02]
anha:  73%|#######2  | 2404/3300 [14:44<34:18,  2.30s/it, lr: 1.0e-04 loss: 2.513e-02]
anha:  73%|#######2  | 2404/3300 [14:46<34:18,  2.30s/it, lr: 1.0e-04 loss: 3.592e-02]
anha:  73%|#######2  | 2404/3300 [14:46<34:18,  2.30s/it, lr: 1.0e-04 loss: 3.592e-02]
anha:  73%|#######2  | 2405/3300 [14:46<31:51,  2.14s/it, lr: 1.0e-04 loss: 3.592e-02]
anha:  73%|#######2  | 2405/3300 [14:46<31:51,  2.14s/it, lr: 1.0e-04 loss: 3.592e-02]
anha:  73%|#######2  | 2405/3300 [14:48<31:51,  2.14s/it, lr: 1.0e-04 loss: 5.476e-02]
anha:  73%|#######2  | 2405/3300 [14:48<31:51,  2.14s/it, lr: 1.0e-04 loss: 5.476e-02]
anha:  73%|#######2  | 2406/3300 [14:48<31:06,  2.09s/it, lr: 1.0e-04 loss: 5.476e-02]
anha:  73%|#######2  | 2406/3300 [14:48<31:06,  2.09s/it, lr: 1.0e-04 loss: 5.476e-02]
anha:  73%|#######2  | 2406/3300 [14:49<31:06,  2.09s/it, lr: 1.0e-04 loss: 5.573e-02]
anha:  73%|#######2  | 2406/3300 [14:49<31:06,  2.09s/it, lr: 1.0e-04 loss: 5.573e-02]
anha:  73%|#######2  | 2407/3300 [14:49<26:25,  1.78s/it, lr: 1.0e-04 loss: 5.573e-02]
anha:  73%|#######2  | 2407/3300 [14:49<26:25,  1.78s/it, lr: 1.0e-04 loss: 5.573e-02]
anha:  73%|#######2  | 2407/3300 [14:51<26:25,  1.78s/it, lr: 1.0e-04 loss: 8.387e-02]
anha:  73%|#######2  | 2407/3300 [14:51<26:25,  1.78s/it, lr: 1.0e-04 loss: 8.387e-02]
anha:  73%|#######2  | 2408/3300 [14:51<27:13,  1.83s/it, lr: 1.0e-04 loss: 8.387e-02]
anha:  73%|#######2  | 2408/3300 [14:51<27:13,  1.83s/it, lr: 1.0e-04 loss: 8.387e-02]
anha:  73%|#######2  | 2408/3300 [14:52<27:13,  1.83s/it, lr: 1.0e-04 loss: 3.164e-02]
anha:  73%|#######2  | 2408/3300 [14:52<27:13,  1.83s/it, lr: 1.0e-04 loss: 3.164e-02]
anha:  73%|#######3  | 2409/3300 [14:52<23:39,  1.59s/it, lr: 1.0e-04 loss: 3.164e-02]
anha:  73%|#######3  | 2409/3300 [14:52<23:39,  1.59s/it, lr: 1.0e-04 loss: 3.164e-02]
anha:  73%|#######3  | 2409/3300 [14:54<23:39,  1.59s/it, lr: 1.0e-04 loss: 8.520e-02]
anha:  73%|#######3  | 2409/3300 [14:54<23:39,  1.59s/it, lr: 1.0e-04 loss: 8.520e-02]
anha:  73%|#######3  | 2410/3300 [14:56<23:37,  1.59s/it, lr: 1.0e-04 loss: 3.473e-02]
anha:  73%|#######3  | 2410/3300 [14:56<23:37,  1.59s/it, lr: 1.0e-04 loss: 3.473e-02]
anha:  73%|#######3  | 2411/3300 [14:56<19:30,  1.32s/it, lr: 1.0e-04 loss: 3.473e-02]
anha:  73%|#######3  | 2411/3300 [14:56<19:30,  1.32s/it, lr: 1.0e-04 loss: 3.473e-02]
anha:  73%|#######3  | 2411/3300 [14:59<19:30,  1.32s/it, lr: 1.0e-04 loss: 1.221e-01]
anha:  73%|#######3  | 2411/3300 [14:59<19:30,  1.32s/it, lr: 1.0e-04 loss: 1.221e-01]
anha:  73%|#######3  | 2412/3300 [14:59<26:44,  1.81s/it, lr: 1.0e-04 loss: 1.221e-01]
anha:  73%|#######3  | 2412/3300 [14:59<26:44,  1.81s/it, lr: 1.0e-04 loss: 1.221e-01]
anha:  73%|#######3  | 2412/3300 [15:03<26:44,  1.81s/it, lr: 1.0e-04 loss: 9.110e-02]
anha:  73%|#######3  | 2412/3300 [15:03<26:44,  1.81s/it, lr: 1.0e-04 loss: 9.110e-02]
anha:  73%|#######3  | 2413/3300 [15:03<32:23,  2.19s/it, lr: 1.0e-04 loss: 9.110e-02]
anha:  73%|#######3  | 2413/3300 [15:03<32:23,  2.19s/it, lr: 1.0e-04 loss: 9.110e-02]
anha:  73%|#######3  | 2413/3300 [15:06<32:23,  2.19s/it, lr: 1.0e-04 loss: 5.610e-02]
anha:  73%|#######3  | 2413/3300 [15:06<32:23,  2.19s/it, lr: 1.0e-04 loss: 5.610e-02]
anha:  73%|#######3  | 2414/3300 [15:06<37:27,  2.54s/it, lr: 1.0e-04 loss: 5.610e-02]
anha:  73%|#######3  | 2414/3300 [15:06<37:27,  2.54s/it, lr: 1.0e-04 loss: 5.610e-02]
anha:  73%|#######3  | 2414/3300 [15:07<37:27,  2.54s/it, lr: 1.0e-04 loss: 2.043e-01]
anha:  73%|#######3  | 2414/3300 [15:07<37:27,  2.54s/it, lr: 1.0e-04 loss: 2.043e-01]
anha:  73%|#######3  | 2415/3300 [15:07<31:07,  2.11s/it, lr: 1.0e-04 loss: 2.043e-01]
anha:  73%|#######3  | 2415/3300 [15:07<31:07,  2.11s/it, lr: 1.0e-04 loss: 2.043e-01]
anha:  73%|#######3  | 2415/3300 [15:11<31:07,  2.11s/it, lr: 1.0e-04 loss: 5.656e-02]
anha:  73%|#######3  | 2415/3300 [15:11<31:07,  2.11s/it, lr: 1.0e-04 loss: 5.656e-02]
anha:  73%|#######3  | 2416/3300 [15:11<36:23,  2.47s/it, lr: 1.0e-04 loss: 5.656e-02]
anha:  73%|#######3  | 2416/3300 [15:11<36:23,  2.47s/it, lr: 1.0e-04 loss: 5.656e-02]
anha:  73%|#######3  | 2416/3300 [15:13<36:23,  2.47s/it, lr: 1.0e-04 loss: 3.613e-02]
anha:  73%|#######3  | 2416/3300 [15:13<36:23,  2.47s/it, lr: 1.0e-04 loss: 3.613e-02]
anha:  73%|#######3  | 2417/3300 [15:13<34:08,  2.32s/it, lr: 1.0e-04 loss: 3.613e-02]
anha:  73%|#######3  | 2417/3300 [15:13<34:08,  2.32s/it, lr: 1.0e-04 loss: 3.613e-02]
anha:  73%|#######3  | 2417/3300 [15:14<34:08,  2.32s/it, lr: 1.0e-04 loss: 6.422e-02]
anha:  73%|#######3  | 2417/3300 [15:14<34:08,  2.32s/it, lr: 1.0e-04 loss: 6.422e-02]
anha:  73%|#######3  | 2418/3300 [15:14<32:31,  2.21s/it, lr: 1.0e-04 loss: 6.422e-02]
anha:  73%|#######3  | 2418/3300 [15:14<32:31,  2.21s/it, lr: 1.0e-04 loss: 6.422e-02]
anha:  73%|#######3  | 2418/3300 [15:15<32:31,  2.21s/it, lr: 1.0e-04 loss: 6.092e-02]
anha:  73%|#######3  | 2418/3300 [15:15<32:31,  2.21s/it, lr: 1.0e-04 loss: 6.092e-02]
anha:  73%|#######3  | 2419/3300 [15:15<27:13,  1.85s/it, lr: 1.0e-04 loss: 6.092e-02]
anha:  73%|#######3  | 2419/3300 [15:15<27:13,  1.85s/it, lr: 1.0e-04 loss: 6.092e-02]
anha:  73%|#######3  | 2419/3300 [15:19<27:13,  1.85s/it, lr: 1.0e-04 loss: 4.599e-02]
anha:  73%|#######3  | 2419/3300 [15:19<27:13,  1.85s/it, lr: 1.0e-04 loss: 4.599e-02]
anha:  73%|#######3  | 2420/3300 [15:21<27:11,  1.85s/it, lr: 1.0e-04 loss: 4.327e-02]
anha:  73%|#######3  | 2420/3300 [15:21<27:11,  1.85s/it, lr: 1.0e-04 loss: 4.327e-02]
anha:  73%|#######3  | 2421/3300 [15:21<21:21,  1.46s/it, lr: 1.0e-04 loss: 4.327e-02]
anha:  73%|#######3  | 2421/3300 [15:21<21:21,  1.46s/it, lr: 1.0e-04 loss: 4.327e-02]
anha:  73%|#######3  | 2421/3300 [15:24<21:21,  1.46s/it, lr: 1.0e-04 loss: 4.434e-02]
anha:  73%|#######3  | 2421/3300 [15:24<21:21,  1.46s/it, lr: 1.0e-04 loss: 4.434e-02]
anha:  73%|#######3  | 2422/3300 [15:24<28:13,  1.93s/it, lr: 1.0e-04 loss: 4.434e-02]
anha:  73%|#######3  | 2422/3300 [15:24<28:13,  1.93s/it, lr: 1.0e-04 loss: 4.434e-02]
anha:  73%|#######3  | 2422/3300 [15:28<28:13,  1.93s/it, lr: 1.0e-04 loss: 1.611e-01]
anha:  73%|#######3  | 2422/3300 [15:28<28:13,  1.93s/it, lr: 1.0e-04 loss: 1.611e-01]
anha:  73%|#######3  | 2423/3300 [15:28<33:27,  2.29s/it, lr: 1.0e-04 loss: 1.611e-01]
anha:  73%|#######3  | 2423/3300 [15:28<33:27,  2.29s/it, lr: 1.0e-04 loss: 1.611e-01]
anha:  73%|#######3  | 2423/3300 [15:29<33:27,  2.29s/it, lr: 1.0e-04 loss: 5.067e-02]
anha:  73%|#######3  | 2423/3300 [15:29<33:27,  2.29s/it, lr: 1.0e-04 loss: 5.067e-02]
anha:  73%|#######3  | 2424/3300 [15:29<28:21,  1.94s/it, lr: 1.0e-04 loss: 5.067e-02]
anha:  73%|#######3  | 2424/3300 [15:29<28:21,  1.94s/it, lr: 1.0e-04 loss: 5.067e-02]
anha:  73%|#######3  | 2424/3300 [15:30<28:21,  1.94s/it, lr: 1.0e-04 loss: 1.255e-01]
anha:  73%|#######3  | 2424/3300 [15:30<28:21,  1.94s/it, lr: 1.0e-04 loss: 1.255e-01]
anha:  73%|#######3  | 2425/3300 [15:30<27:24,  1.88s/it, lr: 1.0e-04 loss: 1.255e-01]
anha:  73%|#######3  | 2425/3300 [15:30<27:24,  1.88s/it, lr: 1.0e-04 loss: 1.255e-01]
anha:  73%|#######3  | 2425/3300 [15:32<27:24,  1.88s/it, lr: 1.0e-04 loss: 3.008e-02]
anha:  73%|#######3  | 2425/3300 [15:32<27:24,  1.88s/it, lr: 1.0e-04 loss: 3.008e-02]
anha:  74%|#######3  | 2426/3300 [15:32<27:40,  1.90s/it, lr: 1.0e-04 loss: 3.008e-02]
anha:  74%|#######3  | 2426/3300 [15:32<27:40,  1.90s/it, lr: 1.0e-04 loss: 3.008e-02]
anha:  74%|#######3  | 2426/3300 [15:34<27:40,  1.90s/it, lr: 1.0e-04 loss: 8.094e-02]
anha:  74%|#######3  | 2426/3300 [15:34<27:40,  1.90s/it, lr: 1.0e-04 loss: 8.094e-02]
anha:  74%|#######3  | 2427/3300 [15:34<27:51,  1.91s/it, lr: 1.0e-04 loss: 8.094e-02]
anha:  74%|#######3  | 2427/3300 [15:34<27:51,  1.91s/it, lr: 1.0e-04 loss: 8.094e-02]
anha:  74%|#######3  | 2427/3300 [15:35<27:51,  1.91s/it, lr: 1.0e-04 loss: 7.079e-02]
anha:  74%|#######3  | 2427/3300 [15:35<27:51,  1.91s/it, lr: 1.0e-04 loss: 7.079e-02]
anha:  74%|#######3  | 2428/3300 [15:35<23:57,  1.65s/it, lr: 1.0e-04 loss: 7.079e-02]
anha:  74%|#######3  | 2428/3300 [15:35<23:57,  1.65s/it, lr: 1.0e-04 loss: 7.079e-02]
anha:  74%|#######3  | 2428/3300 [15:37<23:57,  1.65s/it, lr: 1.0e-04 loss: 6.320e-02]
anha:  74%|#######3  | 2428/3300 [15:37<23:57,  1.65s/it, lr: 1.0e-04 loss: 6.320e-02]
anha:  74%|#######3  | 2429/3300 [15:37<24:16,  1.67s/it, lr: 1.0e-04 loss: 6.320e-02]
anha:  74%|#######3  | 2429/3300 [15:37<24:16,  1.67s/it, lr: 1.0e-04 loss: 6.320e-02]
anha:  74%|#######3  | 2429/3300 [15:39<24:16,  1.67s/it, lr: 1.0e-04 loss: 4.871e-02]
anha:  74%|#######3  | 2429/3300 [15:39<24:16,  1.67s/it, lr: 1.0e-04 loss: 4.871e-02]
anha:  74%|#######3  | 2430/3300 [15:42<24:14,  1.67s/it, lr: 1.0e-04 loss: 4.362e-02]
anha:  74%|#######3  | 2430/3300 [15:42<24:14,  1.67s/it, lr: 1.0e-04 loss: 4.362e-02]
anha:  74%|#######3  | 2431/3300 [15:42<24:40,  1.70s/it, lr: 1.0e-04 loss: 4.362e-02]
anha:  74%|#######3  | 2431/3300 [15:42<24:40,  1.70s/it, lr: 1.0e-04 loss: 4.362e-02]
anha:  74%|#######3  | 2431/3300 [15:45<24:40,  1.70s/it, lr: 1.0e-04 loss: 3.497e-02]
anha:  74%|#######3  | 2431/3300 [15:45<24:40,  1.70s/it, lr: 1.0e-04 loss: 3.497e-02]
anha:  74%|#######3  | 2432/3300 [15:45<27:43,  1.92s/it, lr: 1.0e-04 loss: 3.497e-02]
anha:  74%|#######3  | 2432/3300 [15:45<27:43,  1.92s/it, lr: 1.0e-04 loss: 3.497e-02]
anha:  74%|#######3  | 2432/3300 [15:46<27:43,  1.92s/it, lr: 1.0e-04 loss: 8.721e-02]
anha:  74%|#######3  | 2432/3300 [15:46<27:43,  1.92s/it, lr: 1.0e-04 loss: 8.721e-02]
anha:  74%|#######3  | 2433/3300 [15:46<24:13,  1.68s/it, lr: 1.0e-04 loss: 8.721e-02]
anha:  74%|#######3  | 2433/3300 [15:46<24:13,  1.68s/it, lr: 1.0e-04 loss: 8.721e-02]
anha:  74%|#######3  | 2433/3300 [15:47<24:13,  1.68s/it, lr: 1.0e-04 loss: 3.502e-02]
anha:  74%|#######3  | 2433/3300 [15:47<24:13,  1.68s/it, lr: 1.0e-04 loss: 3.502e-02]
anha:  74%|#######3  | 2434/3300 [15:47<21:35,  1.50s/it, lr: 1.0e-04 loss: 3.502e-02]
anha:  74%|#######3  | 2434/3300 [15:47<21:35,  1.50s/it, lr: 1.0e-04 loss: 3.502e-02]
anha:  74%|#######3  | 2434/3300 [15:48<21:35,  1.50s/it, lr: 1.0e-04 loss: 5.497e-02]
anha:  74%|#######3  | 2434/3300 [15:48<21:35,  1.50s/it, lr: 1.0e-04 loss: 5.497e-02]
anha:  74%|#######3  | 2435/3300 [15:48<22:33,  1.56s/it, lr: 1.0e-04 loss: 5.497e-02]
anha:  74%|#######3  | 2435/3300 [15:48<22:33,  1.56s/it, lr: 1.0e-04 loss: 5.497e-02]
anha:  74%|#######3  | 2435/3300 [15:49<22:33,  1.56s/it, lr: 1.0e-04 loss: 5.492e-02]
anha:  74%|#######3  | 2435/3300 [15:49<22:33,  1.56s/it, lr: 1.0e-04 loss: 5.492e-02]
anha:  74%|#######3  | 2436/3300 [15:49<20:09,  1.40s/it, lr: 1.0e-04 loss: 5.492e-02]
anha:  74%|#######3  | 2436/3300 [15:49<20:09,  1.40s/it, lr: 1.0e-04 loss: 5.492e-02]
anha:  74%|#######3  | 2436/3300 [15:53<20:09,  1.40s/it, lr: 1.0e-04 loss: 3.532e-02]
anha:  74%|#######3  | 2436/3300 [15:53<20:09,  1.40s/it, lr: 1.0e-04 loss: 3.532e-02]
anha:  74%|#######3  | 2437/3300 [15:53<27:47,  1.93s/it, lr: 1.0e-04 loss: 3.532e-02]
anha:  74%|#######3  | 2437/3300 [15:53<27:47,  1.93s/it, lr: 1.0e-04 loss: 3.532e-02]
anha:  74%|#######3  | 2437/3300 [15:56<27:47,  1.93s/it, lr: 1.0e-04 loss: 9.958e-02]
anha:  74%|#######3  | 2437/3300 [15:56<27:47,  1.93s/it, lr: 1.0e-04 loss: 9.958e-02]
anha:  74%|#######3  | 2438/3300 [15:56<33:54,  2.36s/it, lr: 1.0e-04 loss: 9.958e-02]
anha:  74%|#######3  | 2438/3300 [15:56<33:54,  2.36s/it, lr: 1.0e-04 loss: 9.958e-02]
anha:  74%|#######3  | 2438/3300 [15:59<33:54,  2.36s/it, lr: 1.0e-04 loss: 5.699e-02]
anha:  74%|#######3  | 2438/3300 [15:59<33:54,  2.36s/it, lr: 1.0e-04 loss: 5.699e-02]
anha:  74%|#######3  | 2439/3300 [15:59<37:15,  2.60s/it, lr: 1.0e-04 loss: 5.699e-02]
anha:  74%|#######3  | 2439/3300 [15:59<37:15,  2.60s/it, lr: 1.0e-04 loss: 5.699e-02]
anha:  74%|#######3  | 2439/3300 [16:03<37:15,  2.60s/it, lr: 1.0e-04 loss: 5.927e-02]
anha:  74%|#######3  | 2439/3300 [16:03<37:15,  2.60s/it, lr: 1.0e-04 loss: 5.927e-02]
anha:  74%|#######3  | 2440/3300 [16:05<37:12,  2.60s/it, lr: 1.0e-04 loss: 5.275e-02]
anha:  74%|#######3  | 2440/3300 [16:05<37:12,  2.60s/it, lr: 1.0e-04 loss: 5.275e-02]
anha:  74%|#######3  | 2441/3300 [16:05<26:38,  1.86s/it, lr: 1.0e-04 loss: 5.275e-02]
anha:  74%|#######3  | 2441/3300 [16:05<26:38,  1.86s/it, lr: 1.0e-04 loss: 5.275e-02]
anha:  74%|#######3  | 2441/3300 [16:06<26:38,  1.86s/it, lr: 1.0e-04 loss: 5.614e-02]
anha:  74%|#######3  | 2441/3300 [16:06<26:38,  1.86s/it, lr: 1.0e-04 loss: 5.614e-02]
anha:  74%|#######4  | 2442/3300 [16:06<26:32,  1.86s/it, lr: 1.0e-04 loss: 5.614e-02]
anha:  74%|#######4  | 2442/3300 [16:06<26:32,  1.86s/it, lr: 1.0e-04 loss: 5.614e-02]
anha:  74%|#######4  | 2442/3300 [16:08<26:32,  1.86s/it, lr: 1.0e-04 loss: 1.072e-01]
anha:  74%|#######4  | 2442/3300 [16:08<26:32,  1.86s/it, lr: 1.0e-04 loss: 1.072e-01]
anha:  74%|#######4  | 2443/3300 [16:08<23:22,  1.64s/it, lr: 1.0e-04 loss: 1.072e-01]
anha:  74%|#######4  | 2443/3300 [16:08<23:22,  1.64s/it, lr: 1.0e-04 loss: 1.072e-01]
anha:  74%|#######4  | 2443/3300 [16:09<23:22,  1.64s/it, lr: 1.0e-04 loss: 4.402e-02]
anha:  74%|#######4  | 2443/3300 [16:09<23:22,  1.64s/it, lr: 1.0e-04 loss: 4.402e-02]
anha:  74%|#######4  | 2444/3300 [16:09<20:54,  1.47s/it, lr: 1.0e-04 loss: 4.402e-02]
anha:  74%|#######4  | 2444/3300 [16:09<20:54,  1.47s/it, lr: 1.0e-04 loss: 4.402e-02]
anha:  74%|#######4  | 2444/3300 [16:10<20:54,  1.47s/it, lr: 1.0e-04 loss: 2.814e-02]
anha:  74%|#######4  | 2444/3300 [16:10<20:54,  1.47s/it, lr: 1.0e-04 loss: 2.814e-02]
anha:  74%|#######4  | 2445/3300 [16:10<22:49,  1.60s/it, lr: 1.0e-04 loss: 2.814e-02]
anha:  74%|#######4  | 2445/3300 [16:10<22:49,  1.60s/it, lr: 1.0e-04 loss: 2.814e-02]
anha:  74%|#######4  | 2445/3300 [16:11<22:49,  1.60s/it, lr: 1.0e-04 loss: 5.146e-02]
anha:  74%|#######4  | 2445/3300 [16:11<22:49,  1.60s/it, lr: 1.0e-04 loss: 5.146e-02]
anha:  74%|#######4  | 2446/3300 [16:11<20:18,  1.43s/it, lr: 1.0e-04 loss: 5.146e-02]
anha:  74%|#######4  | 2446/3300 [16:11<20:18,  1.43s/it, lr: 1.0e-04 loss: 5.146e-02]
anha:  74%|#######4  | 2446/3300 [16:12<20:18,  1.43s/it, lr: 1.0e-04 loss: 2.577e-01]
anha:  74%|#######4  | 2446/3300 [16:12<20:18,  1.43s/it, lr: 1.0e-04 loss: 2.577e-01]
anha:  74%|#######4  | 2447/3300 [16:12<18:34,  1.31s/it, lr: 1.0e-04 loss: 2.577e-01]
anha:  74%|#######4  | 2447/3300 [16:12<18:34,  1.31s/it, lr: 1.0e-04 loss: 2.577e-01]
anha:  74%|#######4  | 2447/3300 [16:16<18:34,  1.31s/it, lr: 1.0e-04 loss: 6.947e-02]
anha:  74%|#######4  | 2447/3300 [16:16<18:34,  1.31s/it, lr: 1.0e-04 loss: 6.947e-02]
anha:  74%|#######4  | 2448/3300 [16:16<27:25,  1.93s/it, lr: 1.0e-04 loss: 6.947e-02]
anha:  74%|#######4  | 2448/3300 [16:16<27:25,  1.93s/it, lr: 1.0e-04 loss: 6.947e-02]
anha:  74%|#######4  | 2448/3300 [16:19<27:25,  1.93s/it, lr: 1.0e-04 loss: 6.313e-02]
anha:  74%|#######4  | 2448/3300 [16:19<27:25,  1.93s/it, lr: 1.0e-04 loss: 6.313e-02]
anha:  74%|#######4  | 2449/3300 [16:19<33:30,  2.36s/it, lr: 1.0e-04 loss: 6.313e-02]
anha:  74%|#######4  | 2449/3300 [16:19<33:30,  2.36s/it, lr: 1.0e-04 loss: 6.313e-02]
anha:  74%|#######4  | 2449/3300 [16:20<33:30,  2.36s/it, lr: 1.0e-04 loss: 5.959e-02]
anha:  74%|#######4  | 2449/3300 [16:20<33:30,  2.36s/it, lr: 1.0e-04 loss: 5.959e-02]
anha:  74%|#######4  | 2450/3300 [16:22<33:28,  2.36s/it, lr: 1.0e-04 loss: 4.108e-02]
anha:  74%|#######4  | 2450/3300 [16:22<33:28,  2.36s/it, lr: 1.0e-04 loss: 4.108e-02]
anha:  74%|#######4  | 2451/3300 [16:22<24:47,  1.75s/it, lr: 1.0e-04 loss: 4.108e-02]
anha:  74%|#######4  | 2451/3300 [16:22<24:47,  1.75s/it, lr: 1.0e-04 loss: 4.108e-02]
anha:  74%|#######4  | 2451/3300 [16:24<24:47,  1.75s/it, lr: 1.0e-04 loss: 3.587e-02]
anha:  74%|#######4  | 2451/3300 [16:24<24:47,  1.75s/it, lr: 1.0e-04 loss: 3.587e-02]
anha:  74%|#######4  | 2452/3300 [16:24<25:27,  1.80s/it, lr: 1.0e-04 loss: 3.587e-02]
anha:  74%|#######4  | 2452/3300 [16:24<25:27,  1.80s/it, lr: 1.0e-04 loss: 3.587e-02]
anha:  74%|#######4  | 2452/3300 [16:28<25:27,  1.80s/it, lr: 1.0e-04 loss: 5.577e-02]
anha:  74%|#######4  | 2452/3300 [16:28<25:27,  1.80s/it, lr: 1.0e-04 loss: 5.577e-02]
anha:  74%|#######4  | 2453/3300 [16:28<31:27,  2.23s/it, lr: 1.0e-04 loss: 5.577e-02]
anha:  74%|#######4  | 2453/3300 [16:28<31:27,  2.23s/it, lr: 1.0e-04 loss: 5.577e-02]
anha:  74%|#######4  | 2453/3300 [16:29<31:27,  2.23s/it, lr: 1.0e-04 loss: 8.714e-02]
anha:  74%|#######4  | 2453/3300 [16:29<31:27,  2.23s/it, lr: 1.0e-04 loss: 8.714e-02]
anha:  74%|#######4  | 2454/3300 [16:29<26:46,  1.90s/it, lr: 1.0e-04 loss: 8.714e-02]
anha:  74%|#######4  | 2454/3300 [16:29<26:46,  1.90s/it, lr: 1.0e-04 loss: 8.714e-02]
anha:  74%|#######4  | 2454/3300 [16:31<26:46,  1.90s/it, lr: 1.0e-04 loss: 2.052e-01]
anha:  74%|#######4  | 2454/3300 [16:31<26:46,  1.90s/it, lr: 1.0e-04 loss: 2.052e-01]
anha:  74%|#######4  | 2455/3300 [16:31<27:03,  1.92s/it, lr: 1.0e-04 loss: 2.052e-01]
anha:  74%|#######4  | 2455/3300 [16:31<27:03,  1.92s/it, lr: 1.0e-04 loss: 2.052e-01]
anha:  74%|#######4  | 2455/3300 [16:34<27:03,  1.92s/it, lr: 1.0e-04 loss: 1.317e-01]
anha:  74%|#######4  | 2455/3300 [16:34<27:03,  1.92s/it, lr: 1.0e-04 loss: 1.317e-01]
anha:  74%|#######4  | 2456/3300 [16:34<32:29,  2.31s/it, lr: 1.0e-04 loss: 1.317e-01]
anha:  74%|#######4  | 2456/3300 [16:34<32:29,  2.31s/it, lr: 1.0e-04 loss: 1.317e-01]
anha:  74%|#######4  | 2456/3300 [16:36<32:29,  2.31s/it, lr: 1.0e-04 loss: 3.773e-02]
anha:  74%|#######4  | 2456/3300 [16:36<32:29,  2.31s/it, lr: 1.0e-04 loss: 3.773e-02]
anha:  74%|#######4  | 2457/3300 [16:36<31:06,  2.21s/it, lr: 1.0e-04 loss: 3.773e-02]
anha:  74%|#######4  | 2457/3300 [16:36<31:06,  2.21s/it, lr: 1.0e-04 loss: 3.773e-02]
anha:  74%|#######4  | 2457/3300 [16:39<31:06,  2.21s/it, lr: 1.0e-04 loss: 5.371e-02]
anha:  74%|#######4  | 2457/3300 [16:39<31:06,  2.21s/it, lr: 1.0e-04 loss: 5.371e-02]
anha:  74%|#######4  | 2458/3300 [16:39<35:54,  2.56s/it, lr: 1.0e-04 loss: 5.371e-02]
anha:  74%|#######4  | 2458/3300 [16:39<35:54,  2.56s/it, lr: 1.0e-04 loss: 5.371e-02]
anha:  74%|#######4  | 2458/3300 [16:41<35:54,  2.56s/it, lr: 1.0e-04 loss: 6.787e-02]
anha:  74%|#######4  | 2458/3300 [16:41<35:54,  2.56s/it, lr: 1.0e-04 loss: 6.787e-02]
anha:  75%|#######4  | 2459/3300 [16:41<33:21,  2.38s/it, lr: 1.0e-04 loss: 6.787e-02]
anha:  75%|#######4  | 2459/3300 [16:41<33:21,  2.38s/it, lr: 1.0e-04 loss: 6.787e-02]
anha:  75%|#######4  | 2459/3300 [16:42<33:21,  2.38s/it, lr: 1.0e-04 loss: 4.065e-02]
anha:  75%|#######4  | 2459/3300 [16:42<33:21,  2.38s/it, lr: 1.0e-04 loss: 4.065e-02]
anha:  75%|#######4  | 2460/3300 [16:44<33:18,  2.38s/it, lr: 1.0e-04 loss: 5.182e-02]
anha:  75%|#######4  | 2460/3300 [16:44<33:18,  2.38s/it, lr: 1.0e-04 loss: 5.182e-02]
anha:  75%|#######4  | 2461/3300 [16:44<24:17,  1.74s/it, lr: 1.0e-04 loss: 5.182e-02]
anha:  75%|#######4  | 2461/3300 [16:44<24:17,  1.74s/it, lr: 1.0e-04 loss: 5.182e-02]
anha:  75%|#######4  | 2461/3300 [16:48<24:17,  1.74s/it, lr: 1.0e-04 loss: 4.141e-02]
anha:  75%|#######4  | 2461/3300 [16:48<24:17,  1.74s/it, lr: 1.0e-04 loss: 4.141e-02]
anha:  75%|#######4  | 2462/3300 [16:48<29:52,  2.14s/it, lr: 1.0e-04 loss: 4.141e-02]
anha:  75%|#######4  | 2462/3300 [16:48<29:52,  2.14s/it, lr: 1.0e-04 loss: 4.141e-02]
anha:  75%|#######4  | 2462/3300 [16:50<29:52,  2.14s/it, lr: 1.0e-04 loss: 7.963e-02]
anha:  75%|#######4  | 2462/3300 [16:50<29:52,  2.14s/it, lr: 1.0e-04 loss: 7.963e-02]
anha:  75%|#######4  | 2463/3300 [16:50<29:09,  2.09s/it, lr: 1.0e-04 loss: 7.963e-02]
anha:  75%|#######4  | 2463/3300 [16:50<29:09,  2.09s/it, lr: 1.0e-04 loss: 7.963e-02]
anha:  75%|#######4  | 2463/3300 [16:53<29:09,  2.09s/it, lr: 1.0e-04 loss: 5.522e-02]
anha:  75%|#######4  | 2463/3300 [16:53<29:09,  2.09s/it, lr: 1.0e-04 loss: 5.522e-02]
anha:  75%|#######4  | 2464/3300 [16:53<34:21,  2.47s/it, lr: 1.0e-04 loss: 5.522e-02]
anha:  75%|#######4  | 2464/3300 [16:53<34:21,  2.47s/it, lr: 1.0e-04 loss: 5.522e-02]
anha:  75%|#######4  | 2464/3300 [16:54<34:21,  2.47s/it, lr: 1.0e-04 loss: 6.088e-02]
anha:  75%|#######4  | 2464/3300 [16:54<34:21,  2.47s/it, lr: 1.0e-04 loss: 6.088e-02]
anha:  75%|#######4  | 2465/3300 [16:54<28:37,  2.06s/it, lr: 1.0e-04 loss: 6.088e-02]
anha:  75%|#######4  | 2465/3300 [16:54<28:37,  2.06s/it, lr: 1.0e-04 loss: 6.088e-02]
anha:  75%|#######4  | 2465/3300 [16:55<28:37,  2.06s/it, lr: 1.0e-04 loss: 4.797e-02]
anha:  75%|#######4  | 2465/3300 [16:55<28:37,  2.06s/it, lr: 1.0e-04 loss: 4.797e-02]
anha:  75%|#######4  | 2466/3300 [16:55<24:27,  1.76s/it, lr: 1.0e-04 loss: 4.797e-02]
anha:  75%|#######4  | 2466/3300 [16:55<24:27,  1.76s/it, lr: 1.0e-04 loss: 4.797e-02]
anha:  75%|#######4  | 2466/3300 [16:56<24:27,  1.76s/it, lr: 1.0e-04 loss: 6.344e-02]
anha:  75%|#######4  | 2466/3300 [16:56<24:27,  1.76s/it, lr: 1.0e-04 loss: 6.344e-02]
anha:  75%|#######4  | 2467/3300 [16:56<21:23,  1.54s/it, lr: 1.0e-04 loss: 6.344e-02]
anha:  75%|#######4  | 2467/3300 [16:56<21:23,  1.54s/it, lr: 1.0e-04 loss: 6.344e-02]
anha:  75%|#######4  | 2467/3300 [16:57<21:23,  1.54s/it, lr: 1.0e-04 loss: 1.376e-01]
anha:  75%|#######4  | 2467/3300 [16:57<21:23,  1.54s/it, lr: 1.0e-04 loss: 1.376e-01]
anha:  75%|#######4  | 2468/3300 [16:57<19:09,  1.38s/it, lr: 1.0e-04 loss: 1.376e-01]
anha:  75%|#######4  | 2468/3300 [16:57<19:09,  1.38s/it, lr: 1.0e-04 loss: 1.376e-01]
anha:  75%|#######4  | 2468/3300 [17:01<19:09,  1.38s/it, lr: 1.0e-04 loss: 4.676e-02]
anha:  75%|#######4  | 2468/3300 [17:01<19:09,  1.38s/it, lr: 1.0e-04 loss: 4.676e-02]
anha:  75%|#######4  | 2469/3300 [17:01<27:48,  2.01s/it, lr: 1.0e-04 loss: 4.676e-02]
anha:  75%|#######4  | 2469/3300 [17:01<27:48,  2.01s/it, lr: 1.0e-04 loss: 4.676e-02]
anha:  75%|#######4  | 2469/3300 [17:02<27:48,  2.01s/it, lr: 1.0e-04 loss: 5.092e-02]
anha:  75%|#######4  | 2469/3300 [17:02<27:48,  2.01s/it, lr: 1.0e-04 loss: 5.092e-02]
anha:  75%|#######4  | 2470/3300 [17:04<27:46,  2.01s/it, lr: 1.0e-04 loss: 3.496e-02]
anha:  75%|#######4  | 2470/3300 [17:04<27:46,  2.01s/it, lr: 1.0e-04 loss: 3.496e-02]
anha:  75%|#######4  | 2471/3300 [17:04<22:59,  1.66s/it, lr: 1.0e-04 loss: 3.496e-02]
anha:  75%|#######4  | 2471/3300 [17:04<22:59,  1.66s/it, lr: 1.0e-04 loss: 3.496e-02]
anha:  75%|#######4  | 2471/3300 [17:06<22:59,  1.66s/it, lr: 1.0e-04 loss: 3.778e-02]
anha:  75%|#######4  | 2471/3300 [17:06<22:59,  1.66s/it, lr: 1.0e-04 loss: 3.778e-02]
anha:  75%|#######4  | 2472/3300 [17:06<24:01,  1.74s/it, lr: 1.0e-04 loss: 3.778e-02]
anha:  75%|#######4  | 2472/3300 [17:06<24:01,  1.74s/it, lr: 1.0e-04 loss: 3.778e-02]
anha:  75%|#######4  | 2472/3300 [17:07<24:01,  1.74s/it, lr: 1.0e-04 loss: 1.839e-01]
anha:  75%|#######4  | 2472/3300 [17:07<24:01,  1.74s/it, lr: 1.0e-04 loss: 1.839e-01]
anha:  75%|#######4  | 2473/3300 [17:07<21:26,  1.56s/it, lr: 1.0e-04 loss: 1.839e-01]
anha:  75%|#######4  | 2473/3300 [17:07<21:26,  1.56s/it, lr: 1.0e-04 loss: 1.839e-01]
anha:  75%|#######4  | 2473/3300 [17:09<21:26,  1.56s/it, lr: 1.0e-04 loss: 6.934e-02]
anha:  75%|#######4  | 2473/3300 [17:09<21:26,  1.56s/it, lr: 1.0e-04 loss: 6.934e-02]
anha:  75%|#######4  | 2474/3300 [17:09<22:59,  1.67s/it, lr: 1.0e-04 loss: 6.934e-02]
anha:  75%|#######4  | 2474/3300 [17:09<22:59,  1.67s/it, lr: 1.0e-04 loss: 6.934e-02]
anha:  75%|#######4  | 2474/3300 [17:11<22:59,  1.67s/it, lr: 1.0e-04 loss: 2.216e-01]
anha:  75%|#######4  | 2474/3300 [17:11<22:59,  1.67s/it, lr: 1.0e-04 loss: 2.216e-01]
anha:  75%|#######5  | 2475/3300 [17:11<24:14,  1.76s/it, lr: 1.0e-04 loss: 2.216e-01]
anha:  75%|#######5  | 2475/3300 [17:11<24:14,  1.76s/it, lr: 1.0e-04 loss: 2.216e-01]
anha:  75%|#######5  | 2475/3300 [17:15<24:14,  1.76s/it, lr: 1.0e-04 loss: 4.208e-02]
anha:  75%|#######5  | 2475/3300 [17:15<24:14,  1.76s/it, lr: 1.0e-04 loss: 4.208e-02]
anha:  75%|#######5  | 2476/3300 [17:15<30:56,  2.25s/it, lr: 1.0e-04 loss: 4.208e-02]
anha:  75%|#######5  | 2476/3300 [17:15<30:56,  2.25s/it, lr: 1.0e-04 loss: 4.208e-02]
anha:  75%|#######5  | 2476/3300 [17:16<30:56,  2.25s/it, lr: 1.0e-04 loss: 1.436e-01]
anha:  75%|#######5  | 2476/3300 [17:16<30:56,  2.25s/it, lr: 1.0e-04 loss: 1.436e-01]
anha:  75%|#######5  | 2477/3300 [17:16<26:01,  1.90s/it, lr: 1.0e-04 loss: 1.436e-01]
anha:  75%|#######5  | 2477/3300 [17:16<26:01,  1.90s/it, lr: 1.0e-04 loss: 1.436e-01]
anha:  75%|#######5  | 2477/3300 [17:19<26:01,  1.90s/it, lr: 1.0e-04 loss: 3.917e-02]
anha:  75%|#######5  | 2477/3300 [17:19<26:01,  1.90s/it, lr: 1.0e-04 loss: 3.917e-02]
anha:  75%|#######5  | 2478/3300 [17:19<32:00,  2.34s/it, lr: 1.0e-04 loss: 3.917e-02]
anha:  75%|#######5  | 2478/3300 [17:19<32:00,  2.34s/it, lr: 1.0e-04 loss: 3.917e-02]
anha:  75%|#######5  | 2478/3300 [17:21<32:00,  2.34s/it, lr: 1.0e-04 loss: 9.254e-02]
anha:  75%|#######5  | 2478/3300 [17:21<32:00,  2.34s/it, lr: 1.0e-04 loss: 9.254e-02]
anha:  75%|#######5  | 2479/3300 [17:21<29:28,  2.15s/it, lr: 1.0e-04 loss: 9.254e-02]
anha:  75%|#######5  | 2479/3300 [17:21<29:28,  2.15s/it, lr: 1.0e-04 loss: 9.254e-02]
anha:  75%|#######5  | 2479/3300 [17:22<29:28,  2.15s/it, lr: 1.0e-04 loss: 9.637e-02]
anha:  75%|#######5  | 2479/3300 [17:22<29:28,  2.15s/it, lr: 1.0e-04 loss: 9.637e-02]
anha:  75%|#######5  | 2480/3300 [17:24<29:26,  2.15s/it, lr: 1.0e-04 loss: 1.312e-01]
anha:  75%|#######5  | 2480/3300 [17:24<29:26,  2.15s/it, lr: 1.0e-04 loss: 1.312e-01]
anha:  75%|#######5  | 2481/3300 [17:24<21:21,  1.56s/it, lr: 1.0e-04 loss: 1.312e-01]
anha:  75%|#######5  | 2481/3300 [17:24<21:21,  1.56s/it, lr: 1.0e-04 loss: 1.312e-01]
anha:  75%|#######5  | 2481/3300 [17:25<21:21,  1.56s/it, lr: 1.0e-04 loss: 5.583e-02]
anha:  75%|#######5  | 2481/3300 [17:25<21:21,  1.56s/it, lr: 1.0e-04 loss: 5.583e-02]
anha:  75%|#######5  | 2482/3300 [17:25<21:54,  1.61s/it, lr: 1.0e-04 loss: 5.583e-02]
anha:  75%|#######5  | 2482/3300 [17:25<21:54,  1.61s/it, lr: 1.0e-04 loss: 5.583e-02]
anha:  75%|#######5  | 2482/3300 [17:26<21:54,  1.61s/it, lr: 1.0e-04 loss: 3.916e-02]
anha:  75%|#######5  | 2482/3300 [17:26<21:54,  1.61s/it, lr: 1.0e-04 loss: 3.916e-02]
anha:  75%|#######5  | 2483/3300 [17:26<19:45,  1.45s/it, lr: 1.0e-04 loss: 3.916e-02]
anha:  75%|#######5  | 2483/3300 [17:26<19:45,  1.45s/it, lr: 1.0e-04 loss: 3.916e-02]
anha:  75%|#######5  | 2483/3300 [17:30<19:45,  1.45s/it, lr: 1.0e-04 loss: 7.413e-02]
anha:  75%|#######5  | 2483/3300 [17:30<19:45,  1.45s/it, lr: 1.0e-04 loss: 7.413e-02]
anha:  75%|#######5  | 2484/3300 [17:30<26:19,  1.94s/it, lr: 1.0e-04 loss: 7.413e-02]
anha:  75%|#######5  | 2484/3300 [17:30<26:19,  1.94s/it, lr: 1.0e-04 loss: 7.413e-02]
anha:  75%|#######5  | 2484/3300 [17:32<26:19,  1.94s/it, lr: 1.0e-04 loss: 5.152e-02]
anha:  75%|#######5  | 2484/3300 [17:32<26:19,  1.94s/it, lr: 1.0e-04 loss: 5.152e-02]
anha:  75%|#######5  | 2485/3300 [17:32<26:26,  1.95s/it, lr: 1.0e-04 loss: 5.152e-02]
anha:  75%|#######5  | 2485/3300 [17:32<26:26,  1.95s/it, lr: 1.0e-04 loss: 5.152e-02]
anha:  75%|#######5  | 2485/3300 [17:34<26:26,  1.95s/it, lr: 1.0e-04 loss: 6.856e-02]
anha:  75%|#######5  | 2485/3300 [17:34<26:26,  1.95s/it, lr: 1.0e-04 loss: 6.856e-02]
anha:  75%|#######5  | 2486/3300 [17:34<26:30,  1.95s/it, lr: 1.0e-04 loss: 6.856e-02]
anha:  75%|#######5  | 2486/3300 [17:34<26:30,  1.95s/it, lr: 1.0e-04 loss: 6.856e-02]
anha:  75%|#######5  | 2486/3300 [17:35<26:30,  1.95s/it, lr: 1.0e-04 loss: 2.203e-01]
anha:  75%|#######5  | 2486/3300 [17:35<26:30,  1.95s/it, lr: 1.0e-04 loss: 2.203e-01]
anha:  75%|#######5  | 2487/3300 [17:35<25:55,  1.91s/it, lr: 1.0e-04 loss: 2.203e-01]
anha:  75%|#######5  | 2487/3300 [17:35<25:55,  1.91s/it, lr: 1.0e-04 loss: 2.203e-01]
anha:  75%|#######5  | 2487/3300 [17:37<25:55,  1.91s/it, lr: 1.0e-04 loss: 8.612e-02]
anha:  75%|#######5  | 2487/3300 [17:37<25:55,  1.91s/it, lr: 1.0e-04 loss: 8.612e-02]
anha:  75%|#######5  | 2488/3300 [17:37<26:06,  1.93s/it, lr: 1.0e-04 loss: 8.612e-02]
anha:  75%|#######5  | 2488/3300 [17:37<26:06,  1.93s/it, lr: 1.0e-04 loss: 8.612e-02]
anha:  75%|#######5  | 2488/3300 [17:41<26:06,  1.93s/it, lr: 1.0e-04 loss: 4.427e-02]
anha:  75%|#######5  | 2488/3300 [17:41<26:06,  1.93s/it, lr: 1.0e-04 loss: 4.427e-02]
anha:  75%|#######5  | 2489/3300 [17:41<32:20,  2.39s/it, lr: 1.0e-04 loss: 4.427e-02]
anha:  75%|#######5  | 2489/3300 [17:41<32:20,  2.39s/it, lr: 1.0e-04 loss: 4.427e-02]
anha:  75%|#######5  | 2489/3300 [17:44<32:20,  2.39s/it, lr: 1.0e-04 loss: 1.199e-01]
anha:  75%|#######5  | 2489/3300 [17:44<32:20,  2.39s/it, lr: 1.0e-04 loss: 1.199e-01]
anha:  75%|#######5  | 2490/3300 [17:45<32:18,  2.39s/it, lr: 1.0e-04 loss: 5.529e-02]
anha:  75%|#######5  | 2490/3300 [17:45<32:18,  2.39s/it, lr: 1.0e-04 loss: 5.529e-02]
anha:  75%|#######5  | 2491/3300 [17:45<20:37,  1.53s/it, lr: 1.0e-04 loss: 5.529e-02]
anha:  75%|#######5  | 2491/3300 [17:45<20:37,  1.53s/it, lr: 1.0e-04 loss: 5.529e-02]
anha:  75%|#######5  | 2491/3300 [17:49<20:37,  1.53s/it, lr: 1.0e-04 loss: 7.001e-02]
anha:  75%|#######5  | 2491/3300 [17:49<20:37,  1.53s/it, lr: 1.0e-04 loss: 7.001e-02]
anha:  76%|#######5  | 2492/3300 [17:49<27:03,  2.01s/it, lr: 1.0e-04 loss: 7.001e-02]
anha:  76%|#######5  | 2492/3300 [17:49<27:03,  2.01s/it, lr: 1.0e-04 loss: 7.001e-02]
anha:  76%|#######5  | 2492/3300 [17:51<27:03,  2.01s/it, lr: 1.0e-04 loss: 4.684e-02]
anha:  76%|#######5  | 2492/3300 [17:51<27:03,  2.01s/it, lr: 1.0e-04 loss: 4.684e-02]
anha:  76%|#######5  | 2493/3300 [17:51<26:53,  2.00s/it, lr: 1.0e-04 loss: 4.684e-02]
anha:  76%|#######5  | 2493/3300 [17:51<26:53,  2.00s/it, lr: 1.0e-04 loss: 4.684e-02]
anha:  76%|#######5  | 2493/3300 [17:54<26:53,  2.00s/it, lr: 1.0e-04 loss: 4.146e-02]
anha:  76%|#######5  | 2493/3300 [17:54<26:53,  2.00s/it, lr: 1.0e-04 loss: 4.146e-02]
anha:  76%|#######5  | 2494/3300 [17:54<31:54,  2.37s/it, lr: 1.0e-04 loss: 4.146e-02]
anha:  76%|#######5  | 2494/3300 [17:54<31:54,  2.37s/it, lr: 1.0e-04 loss: 4.146e-02]
anha:  76%|#######5  | 2494/3300 [17:55<31:54,  2.37s/it, lr: 1.0e-04 loss: 8.133e-02]
anha:  76%|#######5  | 2494/3300 [17:55<31:54,  2.37s/it, lr: 1.0e-04 loss: 8.133e-02]
anha:  76%|#######5  | 2495/3300 [17:55<26:41,  1.99s/it, lr: 1.0e-04 loss: 8.133e-02]
anha:  76%|#######5  | 2495/3300 [17:55<26:41,  1.99s/it, lr: 1.0e-04 loss: 8.133e-02]
anha:  76%|#######5  | 2495/3300 [17:58<26:41,  1.99s/it, lr: 1.0e-04 loss: 3.782e-02]
anha:  76%|#######5  | 2495/3300 [17:58<26:41,  1.99s/it, lr: 1.0e-04 loss: 3.782e-02]
anha:  76%|#######5  | 2496/3300 [17:58<32:00,  2.39s/it, lr: 1.0e-04 loss: 3.782e-02]
anha:  76%|#######5  | 2496/3300 [17:58<32:00,  2.39s/it, lr: 1.0e-04 loss: 3.782e-02]
anha:  76%|#######5  | 2496/3300 [18:02<32:00,  2.39s/it, lr: 1.0e-04 loss: 3.784e-02]
anha:  76%|#######5  | 2496/3300 [18:02<32:00,  2.39s/it, lr: 1.0e-04 loss: 3.784e-02]
anha:  76%|#######5  | 2497/3300 [18:02<34:57,  2.61s/it, lr: 1.0e-04 loss: 3.784e-02]
anha:  76%|#######5  | 2497/3300 [18:02<34:57,  2.61s/it, lr: 1.0e-04 loss: 3.784e-02]
anha:  76%|#######5  | 2497/3300 [18:03<34:57,  2.61s/it, lr: 1.0e-04 loss: 1.766e-01]
anha:  76%|#######5  | 2497/3300 [18:03<34:57,  2.61s/it, lr: 1.0e-04 loss: 1.766e-01]
anha:  76%|#######5  | 2498/3300 [18:03<28:40,  2.14s/it, lr: 1.0e-04 loss: 1.766e-01]
anha:  76%|#######5  | 2498/3300 [18:03<28:40,  2.14s/it, lr: 1.0e-04 loss: 1.766e-01]
anha:  76%|#######5  | 2498/3300 [18:06<28:40,  2.14s/it, lr: 1.0e-04 loss: 7.639e-02]
anha:  76%|#######5  | 2498/3300 [18:06<28:40,  2.14s/it, lr: 1.0e-04 loss: 7.639e-02]
anha:  76%|#######5  | 2499/3300 [18:06<33:07,  2.48s/it, lr: 1.0e-04 loss: 7.639e-02]
anha:  76%|#######5  | 2499/3300 [18:06<33:07,  2.48s/it, lr: 1.0e-04 loss: 7.639e-02]
anha:  76%|#######5  | 2499/3300 [18:09<33:07,  2.48s/it, lr: 1.0e-04 loss: 3.294e-02]
anha:  76%|#######5  | 2499/3300 [18:09<33:07,  2.48s/it, lr: 1.0e-04 loss: 3.294e-02]


Generating Images: 0it [00:00, ?it/s]
Generating Images: 0it [00:00, ?it/s]


                                     
                                     
anha:  76%|#######5  | 2500/3300 [18:13<33:04,  2.48s/it, lr: 1.0e-04 loss: 5.326e-02]
anha:  76%|#######5  | 2500/3300 [18:13<33:04,  2.48s/it, lr: 1.0e-04 loss: 5.326e-02]
anha:  76%|#######5  | 2501/3300 [18:13<28:31,  2.14s/it, lr: 1.0e-04 loss: 5.326e-02]
anha:  76%|#######5  | 2501/3300 [18:13<28:31,  2.14s/it, lr: 1.0e-04 loss: 5.326e-02]
anha:  76%|#######5  | 2501/3300 [18:16<28:31,  2.14s/it, lr: 1.0e-04 loss: 3.625e-01]
anha:  76%|#######5  | 2501/3300 [18:16<28:31,  2.14s/it, lr: 1.0e-04 loss: 3.625e-01]
anha:  76%|#######5  | 2502/3300 [18:16<32:13,  2.42s/it, lr: 1.0e-04 loss: 3.625e-01]
anha:  76%|#######5  | 2502/3300 [18:16<32:13,  2.42s/it, lr: 1.0e-04 loss: 3.625e-01]
anha:  76%|#######5  | 2502/3300 [18:18<32:13,  2.42s/it, lr: 1.0e-04 loss: 9.813e-02]
anha:  76%|#######5  | 2502/3300 [18:18<32:13,  2.42s/it, lr: 1.0e-04 loss: 9.813e-02]
anha:  76%|#######5  | 2503/3300 [18:18<29:48,  2.24s/it, lr: 1.0e-04 loss: 9.813e-02]
anha:  76%|#######5  | 2503/3300 [18:18<29:48,  2.24s/it, lr: 1.0e-04 loss: 9.813e-02]
anha:  76%|#######5  | 2503/3300 [18:19<29:48,  2.24s/it, lr: 1.0e-04 loss: 3.639e-02]
anha:  76%|#######5  | 2503/3300 [18:19<29:48,  2.24s/it, lr: 1.0e-04 loss: 3.639e-02]
anha:  76%|#######5  | 2504/3300 [18:19<25:18,  1.91s/it, lr: 1.0e-04 loss: 3.639e-02]
anha:  76%|#######5  | 2504/3300 [18:19<25:18,  1.91s/it, lr: 1.0e-04 loss: 3.639e-02]
anha:  76%|#######5  | 2504/3300 [18:20<25:18,  1.91s/it, lr: 1.0e-04 loss: 6.958e-02]
anha:  76%|#######5  | 2504/3300 [18:20<25:18,  1.91s/it, lr: 1.0e-04 loss: 6.958e-02]
anha:  76%|#######5  | 2505/3300 [18:20<21:57,  1.66s/it, lr: 1.0e-04 loss: 6.958e-02]
anha:  76%|#######5  | 2505/3300 [18:20<21:57,  1.66s/it, lr: 1.0e-04 loss: 6.958e-02]
anha:  76%|#######5  | 2505/3300 [18:21<21:57,  1.66s/it, lr: 1.0e-04 loss: 7.733e-02]
anha:  76%|#######5  | 2505/3300 [18:21<21:57,  1.66s/it, lr: 1.0e-04 loss: 7.733e-02]
anha:  76%|#######5  | 2506/3300 [18:21<19:24,  1.47s/it, lr: 1.0e-04 loss: 7.733e-02]
anha:  76%|#######5  | 2506/3300 [18:21<19:24,  1.47s/it, lr: 1.0e-04 loss: 7.733e-02]
anha:  76%|#######5  | 2506/3300 [18:22<19:24,  1.47s/it, lr: 1.0e-04 loss: 1.061e-01]
anha:  76%|#######5  | 2506/3300 [18:22<19:24,  1.47s/it, lr: 1.0e-04 loss: 1.061e-01]
anha:  76%|#######5  | 2507/3300 [18:22<20:30,  1.55s/it, lr: 1.0e-04 loss: 1.061e-01]
anha:  76%|#######5  | 2507/3300 [18:22<20:30,  1.55s/it, lr: 1.0e-04 loss: 1.061e-01]
anha:  76%|#######5  | 2507/3300 [18:24<20:30,  1.55s/it, lr: 1.0e-04 loss: 6.064e-02]
anha:  76%|#######5  | 2507/3300 [18:24<20:30,  1.55s/it, lr: 1.0e-04 loss: 6.064e-02]
anha:  76%|#######6  | 2508/3300 [18:24<21:09,  1.60s/it, lr: 1.0e-04 loss: 6.064e-02]
anha:  76%|#######6  | 2508/3300 [18:24<21:09,  1.60s/it, lr: 1.0e-04 loss: 6.064e-02]
anha:  76%|#######6  | 2508/3300 [18:26<21:09,  1.60s/it, lr: 1.0e-04 loss: 7.974e-02]
anha:  76%|#######6  | 2508/3300 [18:26<21:09,  1.60s/it, lr: 1.0e-04 loss: 7.974e-02]
anha:  76%|#######6  | 2509/3300 [18:26<22:35,  1.71s/it, lr: 1.0e-04 loss: 7.974e-02]
anha:  76%|#######6  | 2509/3300 [18:26<22:35,  1.71s/it, lr: 1.0e-04 loss: 7.974e-02]
anha:  76%|#######6  | 2509/3300 [18:28<22:35,  1.71s/it, lr: 1.0e-04 loss: 4.654e-02]
anha:  76%|#######6  | 2509/3300 [18:28<22:35,  1.71s/it, lr: 1.0e-04 loss: 4.654e-02]
anha:  76%|#######6  | 2510/3300 [18:30<22:33,  1.71s/it, lr: 1.0e-04 loss: 2.285e-01]
anha:  76%|#######6  | 2510/3300 [18:30<22:33,  1.71s/it, lr: 1.0e-04 loss: 2.285e-01]
anha:  76%|#######6  | 2511/3300 [18:30<17:22,  1.32s/it, lr: 1.0e-04 loss: 2.285e-01]
anha:  76%|#######6  | 2511/3300 [18:30<17:22,  1.32s/it, lr: 1.0e-04 loss: 2.285e-01]
anha:  76%|#######6  | 2511/3300 [18:32<17:22,  1.32s/it, lr: 1.0e-04 loss: 3.595e-02]
anha:  76%|#######6  | 2511/3300 [18:32<17:22,  1.32s/it, lr: 1.0e-04 loss: 3.595e-02]
anha:  76%|#######6  | 2512/3300 [18:32<19:29,  1.48s/it, lr: 1.0e-04 loss: 3.595e-02]
anha:  76%|#######6  | 2512/3300 [18:32<19:29,  1.48s/it, lr: 1.0e-04 loss: 3.595e-02]
anha:  76%|#######6  | 2512/3300 [18:33<19:29,  1.48s/it, lr: 1.0e-04 loss: 4.199e-02]
anha:  76%|#######6  | 2512/3300 [18:33<19:29,  1.48s/it, lr: 1.0e-04 loss: 4.199e-02]
anha:  76%|#######6  | 2513/3300 [18:33<17:46,  1.36s/it, lr: 1.0e-04 loss: 4.199e-02]
anha:  76%|#######6  | 2513/3300 [18:33<17:46,  1.36s/it, lr: 1.0e-04 loss: 4.199e-02]
anha:  76%|#######6  | 2513/3300 [18:35<17:46,  1.36s/it, lr: 1.0e-04 loss: 1.227e-01]
anha:  76%|#######6  | 2513/3300 [18:35<17:46,  1.36s/it, lr: 1.0e-04 loss: 1.227e-01]
anha:  76%|#######6  | 2514/3300 [18:35<19:55,  1.52s/it, lr: 1.0e-04 loss: 1.227e-01]
anha:  76%|#######6  | 2514/3300 [18:35<19:55,  1.52s/it, lr: 1.0e-04 loss: 1.227e-01]
anha:  76%|#######6  | 2514/3300 [18:36<19:55,  1.52s/it, lr: 1.0e-04 loss: 5.987e-02]
anha:  76%|#######6  | 2514/3300 [18:36<19:55,  1.52s/it, lr: 1.0e-04 loss: 5.987e-02]
anha:  76%|#######6  | 2515/3300 [18:36<18:02,  1.38s/it, lr: 1.0e-04 loss: 5.987e-02]
anha:  76%|#######6  | 2515/3300 [18:36<18:02,  1.38s/it, lr: 1.0e-04 loss: 5.987e-02]
anha:  76%|#######6  | 2515/3300 [18:38<18:02,  1.38s/it, lr: 1.0e-04 loss: 4.333e-02]
anha:  76%|#######6  | 2515/3300 [18:38<18:02,  1.38s/it, lr: 1.0e-04 loss: 4.333e-02]
anha:  76%|#######6  | 2516/3300 [18:38<19:21,  1.48s/it, lr: 1.0e-04 loss: 4.333e-02]
anha:  76%|#######6  | 2516/3300 [18:38<19:21,  1.48s/it, lr: 1.0e-04 loss: 4.333e-02]
anha:  76%|#######6  | 2516/3300 [18:39<19:21,  1.48s/it, lr: 1.0e-04 loss: 4.576e-02]
anha:  76%|#######6  | 2516/3300 [18:39<19:21,  1.48s/it, lr: 1.0e-04 loss: 4.576e-02]
anha:  76%|#######6  | 2517/3300 [18:39<17:28,  1.34s/it, lr: 1.0e-04 loss: 4.576e-02]
anha:  76%|#######6  | 2517/3300 [18:39<17:28,  1.34s/it, lr: 1.0e-04 loss: 4.576e-02]
anha:  76%|#######6  | 2517/3300 [18:42<17:28,  1.34s/it, lr: 1.0e-04 loss: 5.109e-02]
anha:  76%|#######6  | 2517/3300 [18:42<17:28,  1.34s/it, lr: 1.0e-04 loss: 5.109e-02]
anha:  76%|#######6  | 2518/3300 [18:42<25:37,  1.97s/it, lr: 1.0e-04 loss: 5.109e-02]
anha:  76%|#######6  | 2518/3300 [18:42<25:37,  1.97s/it, lr: 1.0e-04 loss: 5.109e-02]
anha:  76%|#######6  | 2518/3300 [18:43<25:37,  1.97s/it, lr: 1.0e-04 loss: 1.072e-01]
anha:  76%|#######6  | 2518/3300 [18:43<25:37,  1.97s/it, lr: 1.0e-04 loss: 1.072e-01]
anha:  76%|#######6  | 2519/3300 [18:43<22:23,  1.72s/it, lr: 1.0e-04 loss: 1.072e-01]
anha:  76%|#######6  | 2519/3300 [18:43<22:23,  1.72s/it, lr: 1.0e-04 loss: 1.072e-01]
anha:  76%|#######6  | 2519/3300 [18:45<22:23,  1.72s/it, lr: 1.0e-04 loss: 5.105e-02]
anha:  76%|#######6  | 2519/3300 [18:45<22:23,  1.72s/it, lr: 1.0e-04 loss: 5.105e-02]
anha:  76%|#######6  | 2520/3300 [18:46<22:21,  1.72s/it, lr: 1.0e-04 loss: 3.931e-02]
anha:  76%|#######6  | 2520/3300 [18:46<22:21,  1.72s/it, lr: 1.0e-04 loss: 3.931e-02]
anha:  76%|#######6  | 2521/3300 [18:46<15:06,  1.16s/it, lr: 1.0e-04 loss: 3.931e-02]
anha:  76%|#######6  | 2521/3300 [18:46<15:06,  1.16s/it, lr: 1.0e-04 loss: 3.931e-02]
anha:  76%|#######6  | 2521/3300 [18:47<15:06,  1.16s/it, lr: 1.0e-04 loss: 5.175e-02]
anha:  76%|#######6  | 2521/3300 [18:47<15:06,  1.16s/it, lr: 1.0e-04 loss: 5.175e-02]
anha:  76%|#######6  | 2522/3300 [18:47<14:34,  1.12s/it, lr: 1.0e-04 loss: 5.175e-02]
anha:  76%|#######6  | 2522/3300 [18:47<14:34,  1.12s/it, lr: 1.0e-04 loss: 5.175e-02]
anha:  76%|#######6  | 2522/3300 [18:48<14:34,  1.12s/it, lr: 1.0e-04 loss: 1.352e-01]
anha:  76%|#######6  | 2522/3300 [18:48<14:34,  1.12s/it, lr: 1.0e-04 loss: 1.352e-01]
anha:  76%|#######6  | 2523/3300 [18:48<14:07,  1.09s/it, lr: 1.0e-04 loss: 1.352e-01]
anha:  76%|#######6  | 2523/3300 [18:48<14:07,  1.09s/it, lr: 1.0e-04 loss: 1.352e-01]
anha:  76%|#######6  | 2523/3300 [18:51<14:07,  1.09s/it, lr: 1.0e-04 loss: 6.158e-02]
anha:  76%|#######6  | 2523/3300 [18:51<14:07,  1.09s/it, lr: 1.0e-04 loss: 6.158e-02]
anha:  76%|#######6  | 2524/3300 [18:51<22:27,  1.74s/it, lr: 1.0e-04 loss: 6.158e-02]
anha:  76%|#######6  | 2524/3300 [18:51<22:27,  1.74s/it, lr: 1.0e-04 loss: 6.158e-02]
anha:  76%|#######6  | 2524/3300 [18:52<22:27,  1.74s/it, lr: 1.0e-04 loss: 2.279e-01]
anha:  76%|#######6  | 2524/3300 [18:52<22:27,  1.74s/it, lr: 1.0e-04 loss: 2.279e-01]
anha:  77%|#######6  | 2525/3300 [18:52<19:51,  1.54s/it, lr: 1.0e-04 loss: 2.279e-01]
anha:  77%|#######6  | 2525/3300 [18:52<19:51,  1.54s/it, lr: 1.0e-04 loss: 2.279e-01]
anha:  77%|#######6  | 2525/3300 [18:56<19:51,  1.54s/it, lr: 1.0e-04 loss: 4.193e-02]
anha:  77%|#######6  | 2525/3300 [18:56<19:51,  1.54s/it, lr: 1.0e-04 loss: 4.193e-02]
anha:  77%|#######6  | 2526/3300 [18:56<26:35,  2.06s/it, lr: 1.0e-04 loss: 4.193e-02]
anha:  77%|#######6  | 2526/3300 [18:56<26:35,  2.06s/it, lr: 1.0e-04 loss: 4.193e-02]
anha:  77%|#######6  | 2526/3300 [18:59<26:35,  2.06s/it, lr: 1.0e-04 loss: 3.600e-02]
anha:  77%|#######6  | 2526/3300 [18:59<26:35,  2.06s/it, lr: 1.0e-04 loss: 3.600e-02]
anha:  77%|#######6  | 2527/3300 [18:59<31:25,  2.44s/it, lr: 1.0e-04 loss: 3.600e-02]
anha:  77%|#######6  | 2527/3300 [18:59<31:25,  2.44s/it, lr: 1.0e-04 loss: 3.600e-02]
anha:  77%|#######6  | 2527/3300 [19:02<31:25,  2.44s/it, lr: 1.0e-04 loss: 5.550e-02]
anha:  77%|#######6  | 2527/3300 [19:02<31:25,  2.44s/it, lr: 1.0e-04 loss: 5.550e-02]
anha:  77%|#######6  | 2528/3300 [19:02<34:34,  2.69s/it, lr: 1.0e-04 loss: 5.550e-02]
anha:  77%|#######6  | 2528/3300 [19:02<34:34,  2.69s/it, lr: 1.0e-04 loss: 5.550e-02]
anha:  77%|#######6  | 2528/3300 [19:04<34:34,  2.69s/it, lr: 1.0e-04 loss: 1.273e-01]
anha:  77%|#######6  | 2528/3300 [19:04<34:34,  2.69s/it, lr: 1.0e-04 loss: 1.273e-01]
anha:  77%|#######6  | 2529/3300 [19:04<28:32,  2.22s/it, lr: 1.0e-04 loss: 1.273e-01]
anha:  77%|#######6  | 2529/3300 [19:04<28:32,  2.22s/it, lr: 1.0e-04 loss: 1.273e-01]
anha:  77%|#######6  | 2529/3300 [19:06<28:32,  2.22s/it, lr: 1.0e-04 loss: 3.247e-02]
anha:  77%|#######6  | 2529/3300 [19:06<28:32,  2.22s/it, lr: 1.0e-04 loss: 3.247e-02]
anha:  77%|#######6  | 2530/3300 [19:08<28:30,  2.22s/it, lr: 1.0e-04 loss: 6.254e-02]
anha:  77%|#######6  | 2530/3300 [19:08<28:30,  2.22s/it, lr: 1.0e-04 loss: 6.254e-02]
anha:  77%|#######6  | 2531/3300 [19:08<21:15,  1.66s/it, lr: 1.0e-04 loss: 6.254e-02]
anha:  77%|#######6  | 2531/3300 [19:08<21:15,  1.66s/it, lr: 1.0e-04 loss: 6.254e-02]
anha:  77%|#######6  | 2531/3300 [19:09<21:15,  1.66s/it, lr: 1.0e-04 loss: 4.639e-02]
anha:  77%|#######6  | 2531/3300 [19:09<21:15,  1.66s/it, lr: 1.0e-04 loss: 4.639e-02]
anha:  77%|#######6  | 2532/3300 [19:09<19:11,  1.50s/it, lr: 1.0e-04 loss: 4.639e-02]
anha:  77%|#######6  | 2532/3300 [19:09<19:11,  1.50s/it, lr: 1.0e-04 loss: 4.639e-02]
anha:  77%|#######6  | 2532/3300 [19:12<19:11,  1.50s/it, lr: 1.0e-04 loss: 3.725e-02]
anha:  77%|#######6  | 2532/3300 [19:12<19:11,  1.50s/it, lr: 1.0e-04 loss: 3.725e-02]
anha:  77%|#######6  | 2533/3300 [19:12<25:23,  1.99s/it, lr: 1.0e-04 loss: 3.725e-02]
anha:  77%|#######6  | 2533/3300 [19:12<25:23,  1.99s/it, lr: 1.0e-04 loss: 3.725e-02]
anha:  77%|#######6  | 2533/3300 [19:14<25:23,  1.99s/it, lr: 1.0e-04 loss: 6.127e-02]
anha:  77%|#######6  | 2533/3300 [19:14<25:23,  1.99s/it, lr: 1.0e-04 loss: 6.127e-02]
anha:  77%|#######6  | 2534/3300 [19:14<25:21,  1.99s/it, lr: 1.0e-04 loss: 6.127e-02]
anha:  77%|#######6  | 2534/3300 [19:14<25:21,  1.99s/it, lr: 1.0e-04 loss: 6.127e-02]
anha:  77%|#######6  | 2534/3300 [19:16<25:21,  1.99s/it, lr: 1.0e-04 loss: 2.748e-02]
anha:  77%|#######6  | 2534/3300 [19:16<25:21,  1.99s/it, lr: 1.0e-04 loss: 2.748e-02]
anha:  77%|#######6  | 2535/3300 [19:16<25:12,  1.98s/it, lr: 1.0e-04 loss: 2.748e-02]
anha:  77%|#######6  | 2535/3300 [19:16<25:12,  1.98s/it, lr: 1.0e-04 loss: 2.748e-02]
anha:  77%|#######6  | 2535/3300 [19:17<25:12,  1.98s/it, lr: 1.0e-04 loss: 5.636e-02]
anha:  77%|#######6  | 2535/3300 [19:17<25:12,  1.98s/it, lr: 1.0e-04 loss: 5.636e-02]
anha:  77%|#######6  | 2536/3300 [19:17<21:38,  1.70s/it, lr: 1.0e-04 loss: 5.636e-02]
anha:  77%|#######6  | 2536/3300 [19:17<21:38,  1.70s/it, lr: 1.0e-04 loss: 5.636e-02]
anha:  77%|#######6  | 2536/3300 [19:18<21:38,  1.70s/it, lr: 1.0e-04 loss: 3.530e-02]
anha:  77%|#######6  | 2536/3300 [19:18<21:38,  1.70s/it, lr: 1.0e-04 loss: 3.530e-02]
anha:  77%|#######6  | 2537/3300 [19:18<19:03,  1.50s/it, lr: 1.0e-04 loss: 3.530e-02]
anha:  77%|#######6  | 2537/3300 [19:18<19:03,  1.50s/it, lr: 1.0e-04 loss: 3.530e-02]
anha:  77%|#######6  | 2537/3300 [19:21<19:03,  1.50s/it, lr: 1.0e-04 loss: 1.695e-01]
anha:  77%|#######6  | 2537/3300 [19:21<19:03,  1.50s/it, lr: 1.0e-04 loss: 1.695e-01]
anha:  77%|#######6  | 2538/3300 [19:21<26:27,  2.08s/it, lr: 1.0e-04 loss: 1.695e-01]
anha:  77%|#######6  | 2538/3300 [19:21<26:27,  2.08s/it, lr: 1.0e-04 loss: 1.695e-01]
anha:  77%|#######6  | 2538/3300 [19:23<26:27,  2.08s/it, lr: 1.0e-04 loss: 5.201e-02]
anha:  77%|#######6  | 2538/3300 [19:23<26:27,  2.08s/it, lr: 1.0e-04 loss: 5.201e-02]
anha:  77%|#######6  | 2539/3300 [19:23<25:06,  1.98s/it, lr: 1.0e-04 loss: 5.201e-02]
anha:  77%|#######6  | 2539/3300 [19:23<25:06,  1.98s/it, lr: 1.0e-04 loss: 5.201e-02]
anha:  77%|#######6  | 2539/3300 [19:24<25:06,  1.98s/it, lr: 1.0e-04 loss: 3.919e-02]
anha:  77%|#######6  | 2539/3300 [19:24<25:06,  1.98s/it, lr: 1.0e-04 loss: 3.919e-02]
anha:  77%|#######6  | 2540/3300 [19:25<25:04,  1.98s/it, lr: 1.0e-04 loss: 9.331e-02]
anha:  77%|#######6  | 2540/3300 [19:25<25:04,  1.98s/it, lr: 1.0e-04 loss: 9.331e-02]
anha:  77%|#######7  | 2541/3300 [19:25<16:27,  1.30s/it, lr: 1.0e-04 loss: 9.331e-02]
anha:  77%|#######7  | 2541/3300 [19:25<16:27,  1.30s/it, lr: 1.0e-04 loss: 9.331e-02]
anha:  77%|#######7  | 2541/3300 [19:28<16:27,  1.30s/it, lr: 1.0e-04 loss: 7.475e-02]
anha:  77%|#######7  | 2541/3300 [19:28<16:27,  1.30s/it, lr: 1.0e-04 loss: 7.475e-02]
anha:  77%|#######7  | 2542/3300 [19:28<22:27,  1.78s/it, lr: 1.0e-04 loss: 7.475e-02]
anha:  77%|#######7  | 2542/3300 [19:28<22:27,  1.78s/it, lr: 1.0e-04 loss: 7.475e-02]
anha:  77%|#######7  | 2542/3300 [19:31<22:27,  1.78s/it, lr: 1.0e-04 loss: 6.250e-02]
anha:  77%|#######7  | 2542/3300 [19:31<22:27,  1.78s/it, lr: 1.0e-04 loss: 6.250e-02]
anha:  77%|#######7  | 2543/3300 [19:31<24:53,  1.97s/it, lr: 1.0e-04 loss: 6.250e-02]
anha:  77%|#######7  | 2543/3300 [19:31<24:53,  1.97s/it, lr: 1.0e-04 loss: 6.250e-02]
anha:  77%|#######7  | 2543/3300 [19:32<24:53,  1.97s/it, lr: 1.0e-04 loss: 4.910e-02]
anha:  77%|#######7  | 2543/3300 [19:32<24:53,  1.97s/it, lr: 1.0e-04 loss: 4.910e-02]
anha:  77%|#######7  | 2544/3300 [19:32<21:33,  1.71s/it, lr: 1.0e-04 loss: 4.910e-02]
anha:  77%|#######7  | 2544/3300 [19:32<21:33,  1.71s/it, lr: 1.0e-04 loss: 4.910e-02]
anha:  77%|#######7  | 2544/3300 [19:35<21:33,  1.71s/it, lr: 1.0e-04 loss: 5.567e-02]
anha:  77%|#######7  | 2544/3300 [19:35<21:33,  1.71s/it, lr: 1.0e-04 loss: 5.567e-02]
anha:  77%|#######7  | 2545/3300 [19:35<27:21,  2.17s/it, lr: 1.0e-04 loss: 5.567e-02]
anha:  77%|#######7  | 2545/3300 [19:35<27:21,  2.17s/it, lr: 1.0e-04 loss: 5.567e-02]
anha:  77%|#######7  | 2545/3300 [19:36<27:21,  2.17s/it, lr: 1.0e-04 loss: 7.811e-02]
anha:  77%|#######7  | 2545/3300 [19:36<27:21,  2.17s/it, lr: 1.0e-04 loss: 7.811e-02]
anha:  77%|#######7  | 2546/3300 [19:36<23:13,  1.85s/it, lr: 1.0e-04 loss: 7.811e-02]
anha:  77%|#######7  | 2546/3300 [19:36<23:13,  1.85s/it, lr: 1.0e-04 loss: 7.811e-02]
anha:  77%|#######7  | 2546/3300 [19:37<23:13,  1.85s/it, lr: 1.0e-04 loss: 5.709e-02]
anha:  77%|#######7  | 2546/3300 [19:37<23:13,  1.85s/it, lr: 1.0e-04 loss: 5.709e-02]
anha:  77%|#######7  | 2547/3300 [19:37<20:34,  1.64s/it, lr: 1.0e-04 loss: 5.709e-02]
anha:  77%|#######7  | 2547/3300 [19:37<20:34,  1.64s/it, lr: 1.0e-04 loss: 5.709e-02]
anha:  77%|#######7  | 2547/3300 [19:38<20:34,  1.64s/it, lr: 1.0e-04 loss: 8.077e-02]
anha:  77%|#######7  | 2547/3300 [19:38<20:34,  1.64s/it, lr: 1.0e-04 loss: 8.077e-02]
anha:  77%|#######7  | 2548/3300 [19:38<18:10,  1.45s/it, lr: 1.0e-04 loss: 8.077e-02]
anha:  77%|#######7  | 2548/3300 [19:38<18:10,  1.45s/it, lr: 1.0e-04 loss: 8.077e-02]
anha:  77%|#######7  | 2548/3300 [19:42<18:10,  1.45s/it, lr: 1.0e-04 loss: 7.190e-02]
anha:  77%|#######7  | 2548/3300 [19:42<18:10,  1.45s/it, lr: 1.0e-04 loss: 7.190e-02]
anha:  77%|#######7  | 2549/3300 [19:42<25:30,  2.04s/it, lr: 1.0e-04 loss: 7.190e-02]
anha:  77%|#######7  | 2549/3300 [19:42<25:30,  2.04s/it, lr: 1.0e-04 loss: 7.190e-02]
anha:  77%|#######7  | 2549/3300 [19:43<25:30,  2.04s/it, lr: 1.0e-04 loss: 5.074e-02]
anha:  77%|#######7  | 2549/3300 [19:43<25:30,  2.04s/it, lr: 1.0e-04 loss: 5.074e-02]
anha:  77%|#######7  | 2550/3300 [19:45<25:28,  2.04s/it, lr: 1.0e-04 loss: 4.586e-02]
anha:  77%|#######7  | 2550/3300 [19:45<25:28,  2.04s/it, lr: 1.0e-04 loss: 4.586e-02]
anha:  77%|#######7  | 2551/3300 [19:45<19:22,  1.55s/it, lr: 1.0e-04 loss: 4.586e-02]
anha:  77%|#######7  | 2551/3300 [19:45<19:22,  1.55s/it, lr: 1.0e-04 loss: 4.586e-02]
anha:  77%|#######7  | 2551/3300 [19:48<19:22,  1.55s/it, lr: 1.0e-04 loss: 1.447e-01]
anha:  77%|#######7  | 2551/3300 [19:48<19:22,  1.55s/it, lr: 1.0e-04 loss: 1.447e-01]
anha:  77%|#######7  | 2552/3300 [19:48<24:21,  1.95s/it, lr: 1.0e-04 loss: 1.447e-01]
anha:  77%|#######7  | 2552/3300 [19:48<24:21,  1.95s/it, lr: 1.0e-04 loss: 1.447e-01]
anha:  77%|#######7  | 2552/3300 [19:50<24:21,  1.95s/it, lr: 1.0e-04 loss: 8.203e-02]
anha:  77%|#######7  | 2552/3300 [19:50<24:21,  1.95s/it, lr: 1.0e-04 loss: 8.203e-02]
anha:  77%|#######7  | 2553/3300 [19:50<23:33,  1.89s/it, lr: 1.0e-04 loss: 8.203e-02]
anha:  77%|#######7  | 2553/3300 [19:50<23:33,  1.89s/it, lr: 1.0e-04 loss: 8.203e-02]
anha:  77%|#######7  | 2553/3300 [19:53<23:33,  1.89s/it, lr: 1.0e-04 loss: 6.495e-02]
anha:  77%|#######7  | 2553/3300 [19:53<23:33,  1.89s/it, lr: 1.0e-04 loss: 6.495e-02]
anha:  77%|#######7  | 2554/3300 [19:53<28:53,  2.32s/it, lr: 1.0e-04 loss: 6.495e-02]
anha:  77%|#######7  | 2554/3300 [19:53<28:53,  2.32s/it, lr: 1.0e-04 loss: 6.495e-02]
anha:  77%|#######7  | 2554/3300 [19:55<28:53,  2.32s/it, lr: 1.0e-04 loss: 5.998e-02]
anha:  77%|#######7  | 2554/3300 [19:55<28:53,  2.32s/it, lr: 1.0e-04 loss: 5.998e-02]
anha:  77%|#######7  | 2555/3300 [19:55<27:37,  2.23s/it, lr: 1.0e-04 loss: 5.998e-02]
anha:  77%|#######7  | 2555/3300 [19:55<27:37,  2.23s/it, lr: 1.0e-04 loss: 5.998e-02]
anha:  77%|#######7  | 2555/3300 [19:58<27:37,  2.23s/it, lr: 1.0e-04 loss: 5.240e-02]
anha:  77%|#######7  | 2555/3300 [19:58<27:37,  2.23s/it, lr: 1.0e-04 loss: 5.240e-02]
anha:  77%|#######7  | 2556/3300 [19:58<31:20,  2.53s/it, lr: 1.0e-04 loss: 5.240e-02]
anha:  77%|#######7  | 2556/3300 [19:58<31:20,  2.53s/it, lr: 1.0e-04 loss: 5.240e-02]
anha:  77%|#######7  | 2556/3300 [20:02<31:20,  2.53s/it, lr: 1.0e-04 loss: 2.211e-01]
anha:  77%|#######7  | 2556/3300 [20:02<31:20,  2.53s/it, lr: 1.0e-04 loss: 2.211e-01]
anha:  77%|#######7  | 2557/3300 [20:02<34:45,  2.81s/it, lr: 1.0e-04 loss: 2.211e-01]
anha:  77%|#######7  | 2557/3300 [20:02<34:45,  2.81s/it, lr: 1.0e-04 loss: 2.211e-01]
anha:  77%|#######7  | 2557/3300 [20:04<34:45,  2.81s/it, lr: 1.0e-04 loss: 6.319e-02]
anha:  77%|#######7  | 2557/3300 [20:04<34:45,  2.81s/it, lr: 1.0e-04 loss: 6.319e-02]
anha:  78%|#######7  | 2558/3300 [20:04<31:48,  2.57s/it, lr: 1.0e-04 loss: 6.319e-02]
anha:  78%|#######7  | 2558/3300 [20:04<31:48,  2.57s/it, lr: 1.0e-04 loss: 6.319e-02]
anha:  78%|#######7  | 2558/3300 [20:06<31:48,  2.57s/it, lr: 1.0e-04 loss: 4.541e-02]
anha:  78%|#######7  | 2558/3300 [20:06<31:48,  2.57s/it, lr: 1.0e-04 loss: 4.541e-02]
anha:  78%|#######7  | 2559/3300 [20:06<29:30,  2.39s/it, lr: 1.0e-04 loss: 4.541e-02]
anha:  78%|#######7  | 2559/3300 [20:06<29:30,  2.39s/it, lr: 1.0e-04 loss: 4.541e-02]
anha:  78%|#######7  | 2559/3300 [20:08<29:30,  2.39s/it, lr: 1.0e-04 loss: 4.188e-02]
anha:  78%|#######7  | 2559/3300 [20:08<29:30,  2.39s/it, lr: 1.0e-04 loss: 4.188e-02]
anha:  78%|#######7  | 2560/3300 [20:10<29:28,  2.39s/it, lr: 1.0e-04 loss: 4.553e-02]
anha:  78%|#######7  | 2560/3300 [20:10<29:28,  2.39s/it, lr: 1.0e-04 loss: 4.553e-02]
anha:  78%|#######7  | 2561/3300 [20:10<21:28,  1.74s/it, lr: 1.0e-04 loss: 4.553e-02]
anha:  78%|#######7  | 2561/3300 [20:10<21:28,  1.74s/it, lr: 1.0e-04 loss: 4.553e-02]
anha:  78%|#######7  | 2561/3300 [20:13<21:28,  1.74s/it, lr: 1.0e-04 loss: 4.586e-02]
anha:  78%|#######7  | 2561/3300 [20:13<21:28,  1.74s/it, lr: 1.0e-04 loss: 4.586e-02]
anha:  78%|#######7  | 2562/3300 [20:13<26:42,  2.17s/it, lr: 1.0e-04 loss: 4.586e-02]
anha:  78%|#######7  | 2562/3300 [20:13<26:42,  2.17s/it, lr: 1.0e-04 loss: 4.586e-02]
anha:  78%|#######7  | 2562/3300 [20:16<26:42,  2.17s/it, lr: 1.0e-04 loss: 3.038e-02]
anha:  78%|#######7  | 2562/3300 [20:16<26:42,  2.17s/it, lr: 1.0e-04 loss: 3.038e-02]
anha:  78%|#######7  | 2563/3300 [20:16<30:04,  2.45s/it, lr: 1.0e-04 loss: 3.038e-02]
anha:  78%|#######7  | 2563/3300 [20:16<30:04,  2.45s/it, lr: 1.0e-04 loss: 3.038e-02]
anha:  78%|#######7  | 2563/3300 [20:17<30:04,  2.45s/it, lr: 1.0e-04 loss: 2.853e-02]
anha:  78%|#######7  | 2563/3300 [20:17<30:04,  2.45s/it, lr: 1.0e-04 loss: 2.853e-02]
anha:  78%|#######7  | 2564/3300 [20:17<25:18,  2.06s/it, lr: 1.0e-04 loss: 2.853e-02]
anha:  78%|#######7  | 2564/3300 [20:17<25:18,  2.06s/it, lr: 1.0e-04 loss: 2.853e-02]
anha:  78%|#######7  | 2564/3300 [20:21<25:18,  2.06s/it, lr: 1.0e-04 loss: 5.124e-02]
anha:  78%|#######7  | 2564/3300 [20:21<25:18,  2.06s/it, lr: 1.0e-04 loss: 5.124e-02]
anha:  78%|#######7  | 2565/3300 [20:21<29:27,  2.40s/it, lr: 1.0e-04 loss: 5.124e-02]
anha:  78%|#######7  | 2565/3300 [20:21<29:27,  2.40s/it, lr: 1.0e-04 loss: 5.124e-02]
anha:  78%|#######7  | 2565/3300 [20:23<29:27,  2.40s/it, lr: 1.0e-04 loss: 1.002e-01]
anha:  78%|#######7  | 2565/3300 [20:23<29:27,  2.40s/it, lr: 1.0e-04 loss: 1.002e-01]
anha:  78%|#######7  | 2566/3300 [20:23<28:16,  2.31s/it, lr: 1.0e-04 loss: 1.002e-01]
anha:  78%|#######7  | 2566/3300 [20:23<28:16,  2.31s/it, lr: 1.0e-04 loss: 1.002e-01]
anha:  78%|#######7  | 2566/3300 [20:26<28:16,  2.31s/it, lr: 1.0e-04 loss: 5.728e-02]
anha:  78%|#######7  | 2566/3300 [20:26<28:16,  2.31s/it, lr: 1.0e-04 loss: 5.728e-02]
anha:  78%|#######7  | 2567/3300 [20:26<32:04,  2.63s/it, lr: 1.0e-04 loss: 5.728e-02]
anha:  78%|#######7  | 2567/3300 [20:26<32:04,  2.63s/it, lr: 1.0e-04 loss: 5.728e-02]
anha:  78%|#######7  | 2567/3300 [20:29<32:04,  2.63s/it, lr: 1.0e-04 loss: 3.261e-02]
anha:  78%|#######7  | 2567/3300 [20:29<32:04,  2.63s/it, lr: 1.0e-04 loss: 3.261e-02]
anha:  78%|#######7  | 2568/3300 [20:29<34:46,  2.85s/it, lr: 1.0e-04 loss: 3.261e-02]
anha:  78%|#######7  | 2568/3300 [20:29<34:46,  2.85s/it, lr: 1.0e-04 loss: 3.261e-02]
anha:  78%|#######7  | 2568/3300 [20:31<34:46,  2.85s/it, lr: 1.0e-04 loss: 9.859e-02]
anha:  78%|#######7  | 2568/3300 [20:31<34:46,  2.85s/it, lr: 1.0e-04 loss: 9.859e-02]
anha:  78%|#######7  | 2569/3300 [20:31<31:36,  2.59s/it, lr: 1.0e-04 loss: 9.859e-02]
anha:  78%|#######7  | 2569/3300 [20:31<31:36,  2.59s/it, lr: 1.0e-04 loss: 9.859e-02]
anha:  78%|#######7  | 2569/3300 [20:32<31:36,  2.59s/it, lr: 1.0e-04 loss: 9.809e-02]
anha:  78%|#######7  | 2569/3300 [20:32<31:36,  2.59s/it, lr: 1.0e-04 loss: 9.809e-02]
anha:  78%|#######7  | 2570/3300 [20:36<31:33,  2.59s/it, lr: 1.0e-04 loss: 1.208e-01]
anha:  78%|#######7  | 2570/3300 [20:36<31:33,  2.59s/it, lr: 1.0e-04 loss: 1.208e-01]
anha:  78%|#######7  | 2571/3300 [20:36<26:40,  2.20s/it, lr: 1.0e-04 loss: 1.208e-01]
anha:  78%|#######7  | 2571/3300 [20:36<26:40,  2.20s/it, lr: 1.0e-04 loss: 1.208e-01]
anha:  78%|#######7  | 2571/3300 [20:39<26:40,  2.20s/it, lr: 1.0e-04 loss: 4.135e-02]
anha:  78%|#######7  | 2571/3300 [20:39<26:40,  2.20s/it, lr: 1.0e-04 loss: 4.135e-02]
anha:  78%|#######7  | 2572/3300 [20:39<30:07,  2.48s/it, lr: 1.0e-04 loss: 4.135e-02]
anha:  78%|#######7  | 2572/3300 [20:39<30:07,  2.48s/it, lr: 1.0e-04 loss: 4.135e-02]
anha:  78%|#######7  | 2572/3300 [20:40<30:07,  2.48s/it, lr: 1.0e-04 loss: 5.172e-02]
anha:  78%|#######7  | 2572/3300 [20:40<30:07,  2.48s/it, lr: 1.0e-04 loss: 5.172e-02]
anha:  78%|#######7  | 2573/3300 [20:40<25:29,  2.10s/it, lr: 1.0e-04 loss: 5.172e-02]
anha:  78%|#######7  | 2573/3300 [20:40<25:29,  2.10s/it, lr: 1.0e-04 loss: 5.172e-02]
anha:  78%|#######7  | 2573/3300 [20:41<25:29,  2.10s/it, lr: 1.0e-04 loss: 3.703e-02]
anha:  78%|#######7  | 2573/3300 [20:41<25:29,  2.10s/it, lr: 1.0e-04 loss: 3.703e-02]
anha:  78%|#######8  | 2574/3300 [20:41<21:52,  1.81s/it, lr: 1.0e-04 loss: 3.703e-02]
anha:  78%|#######8  | 2574/3300 [20:41<21:52,  1.81s/it, lr: 1.0e-04 loss: 3.703e-02]
anha:  78%|#######8  | 2574/3300 [20:43<21:52,  1.81s/it, lr: 1.0e-04 loss: 3.472e-02]
anha:  78%|#######8  | 2574/3300 [20:43<21:52,  1.81s/it, lr: 1.0e-04 loss: 3.472e-02]
anha:  78%|#######8  | 2575/3300 [20:43<21:31,  1.78s/it, lr: 1.0e-04 loss: 3.472e-02]
anha:  78%|#######8  | 2575/3300 [20:43<21:31,  1.78s/it, lr: 1.0e-04 loss: 3.472e-02]
anha:  78%|#######8  | 2575/3300 [20:44<21:31,  1.78s/it, lr: 1.0e-04 loss: 1.200e-01]
anha:  78%|#######8  | 2575/3300 [20:44<21:31,  1.78s/it, lr: 1.0e-04 loss: 1.200e-01]
anha:  78%|#######8  | 2576/3300 [20:44<19:21,  1.60s/it, lr: 1.0e-04 loss: 1.200e-01]
anha:  78%|#######8  | 2576/3300 [20:44<19:21,  1.60s/it, lr: 1.0e-04 loss: 1.200e-01]
anha:  78%|#######8  | 2576/3300 [20:46<19:21,  1.60s/it, lr: 1.0e-04 loss: 3.996e-02]
anha:  78%|#######8  | 2576/3300 [20:46<19:21,  1.60s/it, lr: 1.0e-04 loss: 3.996e-02]
anha:  78%|#######8  | 2577/3300 [20:46<20:32,  1.70s/it, lr: 1.0e-04 loss: 3.996e-02]
anha:  78%|#######8  | 2577/3300 [20:46<20:32,  1.70s/it, lr: 1.0e-04 loss: 3.996e-02]
anha:  78%|#######8  | 2577/3300 [20:47<20:32,  1.70s/it, lr: 1.0e-04 loss: 7.907e-02]
anha:  78%|#######8  | 2577/3300 [20:47<20:32,  1.70s/it, lr: 1.0e-04 loss: 7.907e-02]
anha:  78%|#######8  | 2578/3300 [20:47<18:07,  1.51s/it, lr: 1.0e-04 loss: 7.907e-02]
anha:  78%|#######8  | 2578/3300 [20:47<18:07,  1.51s/it, lr: 1.0e-04 loss: 7.907e-02]
anha:  78%|#######8  | 2578/3300 [20:48<18:07,  1.51s/it, lr: 1.0e-04 loss: 4.006e-02]
anha:  78%|#######8  | 2578/3300 [20:48<18:07,  1.51s/it, lr: 1.0e-04 loss: 4.006e-02]
anha:  78%|#######8  | 2579/3300 [20:48<16:19,  1.36s/it, lr: 1.0e-04 loss: 4.006e-02]
anha:  78%|#######8  | 2579/3300 [20:48<16:19,  1.36s/it, lr: 1.0e-04 loss: 4.006e-02]
anha:  78%|#######8  | 2579/3300 [20:51<16:19,  1.36s/it, lr: 1.0e-04 loss: 7.118e-02]
anha:  78%|#######8  | 2579/3300 [20:51<16:19,  1.36s/it, lr: 1.0e-04 loss: 7.118e-02]
anha:  78%|#######8  | 2580/3300 [20:52<16:18,  1.36s/it, lr: 1.0e-04 loss: 7.939e-02]
anha:  78%|#######8  | 2580/3300 [20:52<16:18,  1.36s/it, lr: 1.0e-04 loss: 7.939e-02]
anha:  78%|#######8  | 2581/3300 [20:52<11:33,  1.04it/s, lr: 1.0e-04 loss: 7.939e-02]
anha:  78%|#######8  | 2581/3300 [20:52<11:33,  1.04it/s, lr: 1.0e-04 loss: 7.939e-02]
anha:  78%|#######8  | 2581/3300 [20:56<11:33,  1.04it/s, lr: 1.0e-04 loss: 2.606e-02]
anha:  78%|#######8  | 2581/3300 [20:56<11:33,  1.04it/s, lr: 1.0e-04 loss: 2.606e-02]
anha:  78%|#######8  | 2582/3300 [20:56<18:37,  1.56s/it, lr: 1.0e-04 loss: 2.606e-02]
anha:  78%|#######8  | 2582/3300 [20:56<18:37,  1.56s/it, lr: 1.0e-04 loss: 2.606e-02]
anha:  78%|#######8  | 2582/3300 [20:59<18:37,  1.56s/it, lr: 1.0e-04 loss: 7.357e-02]
anha:  78%|#######8  | 2582/3300 [20:59<18:37,  1.56s/it, lr: 1.0e-04 loss: 7.357e-02]
anha:  78%|#######8  | 2583/3300 [20:59<23:58,  2.01s/it, lr: 1.0e-04 loss: 7.357e-02]
anha:  78%|#######8  | 2583/3300 [20:59<23:58,  2.01s/it, lr: 1.0e-04 loss: 7.357e-02]
anha:  78%|#######8  | 2583/3300 [21:02<23:58,  2.01s/it, lr: 1.0e-04 loss: 5.471e-02]
anha:  78%|#######8  | 2583/3300 [21:02<23:58,  2.01s/it, lr: 1.0e-04 loss: 5.471e-02]
anha:  78%|#######8  | 2584/3300 [21:02<27:59,  2.35s/it, lr: 1.0e-04 loss: 5.471e-02]
anha:  78%|#######8  | 2584/3300 [21:02<27:59,  2.35s/it, lr: 1.0e-04 loss: 5.471e-02]
anha:  78%|#######8  | 2584/3300 [21:03<27:59,  2.35s/it, lr: 1.0e-04 loss: 4.526e-02]
anha:  78%|#######8  | 2584/3300 [21:03<27:59,  2.35s/it, lr: 1.0e-04 loss: 4.526e-02]
anha:  78%|#######8  | 2585/3300 [21:03<23:49,  2.00s/it, lr: 1.0e-04 loss: 4.526e-02]
anha:  78%|#######8  | 2585/3300 [21:03<23:49,  2.00s/it, lr: 1.0e-04 loss: 4.526e-02]
anha:  78%|#######8  | 2585/3300 [21:07<23:49,  2.00s/it, lr: 1.0e-04 loss: 7.353e-02]
anha:  78%|#######8  | 2585/3300 [21:07<23:49,  2.00s/it, lr: 1.0e-04 loss: 7.353e-02]
anha:  78%|#######8  | 2586/3300 [21:07<27:58,  2.35s/it, lr: 1.0e-04 loss: 7.353e-02]
anha:  78%|#######8  | 2586/3300 [21:07<27:58,  2.35s/it, lr: 1.0e-04 loss: 7.353e-02]
anha:  78%|#######8  | 2586/3300 [21:10<27:58,  2.35s/it, lr: 1.0e-04 loss: 4.879e-02]
anha:  78%|#######8  | 2586/3300 [21:10<27:58,  2.35s/it, lr: 1.0e-04 loss: 4.879e-02]
anha:  78%|#######8  | 2587/3300 [21:10<31:27,  2.65s/it, lr: 1.0e-04 loss: 4.879e-02]
anha:  78%|#######8  | 2587/3300 [21:10<31:27,  2.65s/it, lr: 1.0e-04 loss: 4.879e-02]
anha:  78%|#######8  | 2587/3300 [21:12<31:27,  2.65s/it, lr: 1.0e-04 loss: 6.463e-02]
anha:  78%|#######8  | 2587/3300 [21:12<31:27,  2.65s/it, lr: 1.0e-04 loss: 6.463e-02]
anha:  78%|#######8  | 2588/3300 [21:12<29:08,  2.46s/it, lr: 1.0e-04 loss: 6.463e-02]
anha:  78%|#######8  | 2588/3300 [21:12<29:08,  2.46s/it, lr: 1.0e-04 loss: 6.463e-02]
anha:  78%|#######8  | 2588/3300 [21:14<29:08,  2.46s/it, lr: 1.0e-04 loss: 3.171e-02]
anha:  78%|#######8  | 2588/3300 [21:14<29:08,  2.46s/it, lr: 1.0e-04 loss: 3.171e-02]
anha:  78%|#######8  | 2589/3300 [21:14<27:20,  2.31s/it, lr: 1.0e-04 loss: 3.171e-02]
anha:  78%|#######8  | 2589/3300 [21:14<27:20,  2.31s/it, lr: 1.0e-04 loss: 3.171e-02]
anha:  78%|#######8  | 2589/3300 [21:17<27:20,  2.31s/it, lr: 1.0e-04 loss: 4.476e-02]
anha:  78%|#######8  | 2589/3300 [21:17<27:20,  2.31s/it, lr: 1.0e-04 loss: 4.476e-02]
anha:  78%|#######8  | 2590/3300 [21:19<27:18,  2.31s/it, lr: 1.0e-04 loss: 8.123e-02]
anha:  78%|#######8  | 2590/3300 [21:19<27:18,  2.31s/it, lr: 1.0e-04 loss: 8.123e-02]
anha:  79%|#######8  | 2591/3300 [21:19<20:06,  1.70s/it, lr: 1.0e-04 loss: 8.123e-02]
anha:  79%|#######8  | 2591/3300 [21:19<20:06,  1.70s/it, lr: 1.0e-04 loss: 8.123e-02]
anha:  79%|#######8  | 2591/3300 [21:22<20:06,  1.70s/it, lr: 1.0e-04 loss: 3.435e-02]
anha:  79%|#######8  | 2591/3300 [21:22<20:06,  1.70s/it, lr: 1.0e-04 loss: 3.435e-02]
anha:  79%|#######8  | 2592/3300 [21:22<22:26,  1.90s/it, lr: 1.0e-04 loss: 3.435e-02]
anha:  79%|#######8  | 2592/3300 [21:22<22:26,  1.90s/it, lr: 1.0e-04 loss: 3.435e-02]
anha:  79%|#######8  | 2592/3300 [21:24<22:26,  1.90s/it, lr: 1.0e-04 loss: 8.328e-02]
anha:  79%|#######8  | 2592/3300 [21:24<22:26,  1.90s/it, lr: 1.0e-04 loss: 8.328e-02]
anha:  79%|#######8  | 2593/3300 [21:24<21:54,  1.86s/it, lr: 1.0e-04 loss: 8.328e-02]
anha:  79%|#######8  | 2593/3300 [21:24<21:54,  1.86s/it, lr: 1.0e-04 loss: 8.328e-02]
anha:  79%|#######8  | 2593/3300 [21:27<21:54,  1.86s/it, lr: 1.0e-04 loss: 1.543e-01]
anha:  79%|#######8  | 2593/3300 [21:27<21:54,  1.86s/it, lr: 1.0e-04 loss: 1.543e-01]
anha:  79%|#######8  | 2594/3300 [21:27<27:05,  2.30s/it, lr: 1.0e-04 loss: 1.543e-01]
anha:  79%|#######8  | 2594/3300 [21:27<27:05,  2.30s/it, lr: 1.0e-04 loss: 1.543e-01]
anha:  79%|#######8  | 2594/3300 [21:28<27:05,  2.30s/it, lr: 1.0e-04 loss: 5.733e-02]
anha:  79%|#######8  | 2594/3300 [21:28<27:05,  2.30s/it, lr: 1.0e-04 loss: 5.733e-02]
anha:  79%|#######8  | 2595/3300 [21:28<22:49,  1.94s/it, lr: 1.0e-04 loss: 5.733e-02]
anha:  79%|#######8  | 2595/3300 [21:28<22:49,  1.94s/it, lr: 1.0e-04 loss: 5.733e-02]
anha:  79%|#######8  | 2595/3300 [21:32<22:49,  1.94s/it, lr: 1.0e-04 loss: 4.921e-02]
anha:  79%|#######8  | 2595/3300 [21:32<22:49,  1.94s/it, lr: 1.0e-04 loss: 4.921e-02]
anha:  79%|#######8  | 2596/3300 [21:32<27:54,  2.38s/it, lr: 1.0e-04 loss: 4.921e-02]
anha:  79%|#######8  | 2596/3300 [21:32<27:54,  2.38s/it, lr: 1.0e-04 loss: 4.921e-02]
anha:  79%|#######8  | 2596/3300 [21:35<27:54,  2.38s/it, lr: 1.0e-04 loss: 4.270e-02]
anha:  79%|#######8  | 2596/3300 [21:35<27:54,  2.38s/it, lr: 1.0e-04 loss: 4.270e-02]
anha:  79%|#######8  | 2597/3300 [21:35<31:36,  2.70s/it, lr: 1.0e-04 loss: 4.270e-02]
anha:  79%|#######8  | 2597/3300 [21:35<31:36,  2.70s/it, lr: 1.0e-04 loss: 4.270e-02]
anha:  79%|#######8  | 2597/3300 [21:37<31:36,  2.70s/it, lr: 1.0e-04 loss: 5.857e-02]
anha:  79%|#######8  | 2597/3300 [21:37<31:36,  2.70s/it, lr: 1.0e-04 loss: 5.857e-02]
anha:  79%|#######8  | 2598/3300 [21:37<29:07,  2.49s/it, lr: 1.0e-04 loss: 5.857e-02]
anha:  79%|#######8  | 2598/3300 [21:37<29:07,  2.49s/it, lr: 1.0e-04 loss: 5.857e-02]
anha:  79%|#######8  | 2598/3300 [21:39<29:07,  2.49s/it, lr: 1.0e-04 loss: 1.128e-01]
anha:  79%|#######8  | 2598/3300 [21:39<29:07,  2.49s/it, lr: 1.0e-04 loss: 1.128e-01]
anha:  79%|#######8  | 2599/3300 [21:39<27:19,  2.34s/it, lr: 1.0e-04 loss: 1.128e-01]
anha:  79%|#######8  | 2599/3300 [21:39<27:19,  2.34s/it, lr: 1.0e-04 loss: 1.128e-01]
anha:  79%|#######8  | 2599/3300 [21:40<27:19,  2.34s/it, lr: 1.0e-04 loss: 3.230e-02]
anha:  79%|#######8  | 2599/3300 [21:40<27:19,  2.34s/it, lr: 1.0e-04 loss: 3.230e-02]
anha:  79%|#######8  | 2600/3300 [21:42<27:17,  2.34s/it, lr: 1.0e-04 loss: 3.432e-02]
anha:  79%|#######8  | 2600/3300 [21:42<27:17,  2.34s/it, lr: 1.0e-04 loss: 3.432e-02]
anha:  79%|#######8  | 2601/3300 [21:42<19:59,  1.72s/it, lr: 1.0e-04 loss: 3.432e-02]
anha:  79%|#######8  | 2601/3300 [21:42<19:59,  1.72s/it, lr: 1.0e-04 loss: 3.432e-02]
anha:  79%|#######8  | 2601/3300 [21:43<19:59,  1.72s/it, lr: 1.0e-04 loss: 3.826e-02]
anha:  79%|#######8  | 2601/3300 [21:43<19:59,  1.72s/it, lr: 1.0e-04 loss: 3.826e-02]
anha:  79%|#######8  | 2602/3300 [21:43<17:53,  1.54s/it, lr: 1.0e-04 loss: 3.826e-02]
anha:  79%|#######8  | 2602/3300 [21:43<17:53,  1.54s/it, lr: 1.0e-04 loss: 3.826e-02]
anha:  79%|#######8  | 2602/3300 [21:46<17:53,  1.54s/it, lr: 1.0e-04 loss: 1.483e-01]
anha:  79%|#######8  | 2602/3300 [21:46<17:53,  1.54s/it, lr: 1.0e-04 loss: 1.483e-01]
anha:  79%|#######8  | 2603/3300 [21:46<23:14,  2.00s/it, lr: 1.0e-04 loss: 1.483e-01]
anha:  79%|#######8  | 2603/3300 [21:46<23:14,  2.00s/it, lr: 1.0e-04 loss: 1.483e-01]
anha:  79%|#######8  | 2603/3300 [21:48<23:14,  2.00s/it, lr: 1.0e-04 loss: 1.234e-01]
anha:  79%|#######8  | 2603/3300 [21:48<23:14,  2.00s/it, lr: 1.0e-04 loss: 1.234e-01]
anha:  79%|#######8  | 2604/3300 [21:48<22:19,  1.92s/it, lr: 1.0e-04 loss: 1.234e-01]
anha:  79%|#######8  | 2604/3300 [21:48<22:19,  1.92s/it, lr: 1.0e-04 loss: 1.234e-01]
anha:  79%|#######8  | 2604/3300 [21:49<22:19,  1.92s/it, lr: 1.0e-04 loss: 9.250e-02]
anha:  79%|#######8  | 2604/3300 [21:49<22:19,  1.92s/it, lr: 1.0e-04 loss: 9.250e-02]
anha:  79%|#######8  | 2605/3300 [21:49<19:19,  1.67s/it, lr: 1.0e-04 loss: 9.250e-02]
anha:  79%|#######8  | 2605/3300 [21:49<19:19,  1.67s/it, lr: 1.0e-04 loss: 9.250e-02]
anha:  79%|#######8  | 2605/3300 [21:50<19:19,  1.67s/it, lr: 1.0e-04 loss: 4.137e-02]
anha:  79%|#######8  | 2605/3300 [21:50<19:19,  1.67s/it, lr: 1.0e-04 loss: 4.137e-02]
anha:  79%|#######8  | 2606/3300 [21:50<17:06,  1.48s/it, lr: 1.0e-04 loss: 4.137e-02]
anha:  79%|#######8  | 2606/3300 [21:50<17:06,  1.48s/it, lr: 1.0e-04 loss: 4.137e-02]
anha:  79%|#######8  | 2606/3300 [21:52<17:06,  1.48s/it, lr: 1.0e-04 loss: 4.523e-02]
anha:  79%|#######8  | 2606/3300 [21:52<17:06,  1.48s/it, lr: 1.0e-04 loss: 4.523e-02]
anha:  79%|#######9  | 2607/3300 [21:52<18:45,  1.62s/it, lr: 1.0e-04 loss: 4.523e-02]
anha:  79%|#######9  | 2607/3300 [21:52<18:45,  1.62s/it, lr: 1.0e-04 loss: 4.523e-02]
anha:  79%|#######9  | 2607/3300 [21:54<18:45,  1.62s/it, lr: 1.0e-04 loss: 2.128e-01]
anha:  79%|#######9  | 2607/3300 [21:54<18:45,  1.62s/it, lr: 1.0e-04 loss: 2.128e-01]
anha:  79%|#######9  | 2608/3300 [21:54<19:05,  1.66s/it, lr: 1.0e-04 loss: 2.128e-01]
anha:  79%|#######9  | 2608/3300 [21:54<19:05,  1.66s/it, lr: 1.0e-04 loss: 2.128e-01]
anha:  79%|#######9  | 2608/3300 [21:57<19:05,  1.66s/it, lr: 1.0e-04 loss: 3.251e-02]
anha:  79%|#######9  | 2608/3300 [21:57<19:05,  1.66s/it, lr: 1.0e-04 loss: 3.251e-02]
anha:  79%|#######9  | 2609/3300 [21:57<24:53,  2.16s/it, lr: 1.0e-04 loss: 3.251e-02]
anha:  79%|#######9  | 2609/3300 [21:57<24:53,  2.16s/it, lr: 1.0e-04 loss: 3.251e-02]
anha:  79%|#######9  | 2609/3300 [21:58<24:53,  2.16s/it, lr: 1.0e-04 loss: 4.862e-02]
anha:  79%|#######9  | 2609/3300 [21:58<24:53,  2.16s/it, lr: 1.0e-04 loss: 4.862e-02]
anha:  79%|#######9  | 2610/3300 [22:02<24:50,  2.16s/it, lr: 1.0e-04 loss: 8.598e-02]
anha:  79%|#######9  | 2610/3300 [22:02<24:50,  2.16s/it, lr: 1.0e-04 loss: 8.598e-02]
anha:  79%|#######9  | 2611/3300 [22:02<22:36,  1.97s/it, lr: 1.0e-04 loss: 8.598e-02]
anha:  79%|#######9  | 2611/3300 [22:02<22:36,  1.97s/it, lr: 1.0e-04 loss: 8.598e-02]
anha:  79%|#######9  | 2611/3300 [22:04<22:36,  1.97s/it, lr: 1.0e-04 loss: 6.720e-02]
anha:  79%|#######9  | 2611/3300 [22:04<22:36,  1.97s/it, lr: 1.0e-04 loss: 6.720e-02]
anha:  79%|#######9  | 2612/3300 [22:04<22:51,  1.99s/it, lr: 1.0e-04 loss: 6.720e-02]
anha:  79%|#######9  | 2612/3300 [22:04<22:51,  1.99s/it, lr: 1.0e-04 loss: 6.720e-02]
anha:  79%|#######9  | 2612/3300 [22:05<22:51,  1.99s/it, lr: 1.0e-04 loss: 8.759e-02]
anha:  79%|#######9  | 2612/3300 [22:05<22:51,  1.99s/it, lr: 1.0e-04 loss: 8.759e-02]
anha:  79%|#######9  | 2613/3300 [22:05<19:52,  1.74s/it, lr: 1.0e-04 loss: 8.759e-02]
anha:  79%|#######9  | 2613/3300 [22:05<19:52,  1.74s/it, lr: 1.0e-04 loss: 8.759e-02]
anha:  79%|#######9  | 2613/3300 [22:06<19:52,  1.74s/it, lr: 1.0e-04 loss: 4.274e-02]
anha:  79%|#######9  | 2613/3300 [22:06<19:52,  1.74s/it, lr: 1.0e-04 loss: 4.274e-02]
anha:  79%|#######9  | 2614/3300 [22:06<19:52,  1.74s/it, lr: 1.0e-04 loss: 4.274e-02]
anha:  79%|#######9  | 2614/3300 [22:06<19:52,  1.74s/it, lr: 1.0e-04 loss: 4.274e-02]
anha:  79%|#######9  | 2614/3300 [22:08<19:52,  1.74s/it, lr: 1.0e-04 loss: 8.089e-02]
anha:  79%|#######9  | 2614/3300 [22:08<19:52,  1.74s/it, lr: 1.0e-04 loss: 8.089e-02]
anha:  79%|#######9  | 2615/3300 [22:08<20:31,  1.80s/it, lr: 1.0e-04 loss: 8.089e-02]
anha:  79%|#######9  | 2615/3300 [22:08<20:31,  1.80s/it, lr: 1.0e-04 loss: 8.089e-02]
anha:  79%|#######9  | 2615/3300 [22:09<20:31,  1.80s/it, lr: 1.0e-04 loss: 7.533e-02]
anha:  79%|#######9  | 2615/3300 [22:09<20:31,  1.80s/it, lr: 1.0e-04 loss: 7.533e-02]
anha:  79%|#######9  | 2616/3300 [22:09<17:59,  1.58s/it, lr: 1.0e-04 loss: 7.533e-02]
anha:  79%|#######9  | 2616/3300 [22:09<17:59,  1.58s/it, lr: 1.0e-04 loss: 7.533e-02]
anha:  79%|#######9  | 2616/3300 [22:11<17:59,  1.58s/it, lr: 1.0e-04 loss: 1.643e-01]
anha:  79%|#######9  | 2616/3300 [22:11<17:59,  1.58s/it, lr: 1.0e-04 loss: 1.643e-01]
anha:  79%|#######9  | 2617/3300 [22:11<19:12,  1.69s/it, lr: 1.0e-04 loss: 1.643e-01]
anha:  79%|#######9  | 2617/3300 [22:11<19:12,  1.69s/it, lr: 1.0e-04 loss: 1.643e-01]
anha:  79%|#######9  | 2617/3300 [22:12<19:12,  1.69s/it, lr: 1.0e-04 loss: 4.268e-02]
anha:  79%|#######9  | 2617/3300 [22:12<19:12,  1.69s/it, lr: 1.0e-04 loss: 4.268e-02]
anha:  79%|#######9  | 2618/3300 [22:12<16:51,  1.48s/it, lr: 1.0e-04 loss: 4.268e-02]
anha:  79%|#######9  | 2618/3300 [22:12<16:51,  1.48s/it, lr: 1.0e-04 loss: 4.268e-02]
anha:  79%|#######9  | 2618/3300 [22:16<16:51,  1.48s/it, lr: 1.0e-04 loss: 2.712e-02]
anha:  79%|#######9  | 2618/3300 [22:16<16:51,  1.48s/it, lr: 1.0e-04 loss: 2.712e-02]
anha:  79%|#######9  | 2619/3300 [22:16<23:08,  2.04s/it, lr: 1.0e-04 loss: 2.712e-02]
anha:  79%|#######9  | 2619/3300 [22:16<23:08,  2.04s/it, lr: 1.0e-04 loss: 2.712e-02]
anha:  79%|#######9  | 2619/3300 [22:17<23:08,  2.04s/it, lr: 1.0e-04 loss: 4.475e-02]
anha:  79%|#######9  | 2619/3300 [22:17<23:08,  2.04s/it, lr: 1.0e-04 loss: 4.475e-02]
anha:  79%|#######9  | 2620/3300 [22:18<23:06,  2.04s/it, lr: 1.0e-04 loss: 1.779e-01]
anha:  79%|#######9  | 2620/3300 [22:18<23:06,  2.04s/it, lr: 1.0e-04 loss: 1.779e-01]
anha:  79%|#######9  | 2621/3300 [22:18<15:22,  1.36s/it, lr: 1.0e-04 loss: 1.779e-01]
anha:  79%|#######9  | 2621/3300 [22:18<15:22,  1.36s/it, lr: 1.0e-04 loss: 1.779e-01]
anha:  79%|#######9  | 2621/3300 [22:21<15:22,  1.36s/it, lr: 1.0e-04 loss: 2.117e-01]
anha:  79%|#######9  | 2621/3300 [22:21<15:22,  1.36s/it, lr: 1.0e-04 loss: 2.117e-01]
anha:  79%|#######9  | 2622/3300 [22:21<20:42,  1.83s/it, lr: 1.0e-04 loss: 2.117e-01]
anha:  79%|#######9  | 2622/3300 [22:21<20:42,  1.83s/it, lr: 1.0e-04 loss: 2.117e-01]
anha:  79%|#######9  | 2622/3300 [22:23<20:42,  1.83s/it, lr: 1.0e-04 loss: 5.605e-02]
anha:  79%|#######9  | 2622/3300 [22:23<20:42,  1.83s/it, lr: 1.0e-04 loss: 5.605e-02]
anha:  79%|#######9  | 2623/3300 [22:23<20:22,  1.81s/it, lr: 1.0e-04 loss: 5.605e-02]
anha:  79%|#######9  | 2623/3300 [22:23<20:22,  1.81s/it, lr: 1.0e-04 loss: 5.605e-02]
anha:  79%|#######9  | 2623/3300 [22:24<20:22,  1.81s/it, lr: 1.0e-04 loss: 7.248e-02]
anha:  79%|#######9  | 2623/3300 [22:24<20:22,  1.81s/it, lr: 1.0e-04 loss: 7.248e-02]
anha:  80%|#######9  | 2624/3300 [22:24<17:53,  1.59s/it, lr: 1.0e-04 loss: 7.248e-02]
anha:  80%|#######9  | 2624/3300 [22:24<17:53,  1.59s/it, lr: 1.0e-04 loss: 7.248e-02]
anha:  80%|#######9  | 2624/3300 [22:26<17:53,  1.59s/it, lr: 1.0e-04 loss: 6.891e-02]
anha:  80%|#######9  | 2624/3300 [22:26<17:53,  1.59s/it, lr: 1.0e-04 loss: 6.891e-02]
anha:  80%|#######9  | 2625/3300 [22:26<18:19,  1.63s/it, lr: 1.0e-04 loss: 6.891e-02]
anha:  80%|#######9  | 2625/3300 [22:26<18:19,  1.63s/it, lr: 1.0e-04 loss: 6.891e-02]
anha:  80%|#######9  | 2625/3300 [22:29<18:19,  1.63s/it, lr: 1.0e-04 loss: 2.364e-01]
anha:  80%|#######9  | 2625/3300 [22:29<18:19,  1.63s/it, lr: 1.0e-04 loss: 2.364e-01]
anha:  80%|#######9  | 2626/3300 [22:29<24:13,  2.16s/it, lr: 1.0e-04 loss: 2.364e-01]
anha:  80%|#######9  | 2626/3300 [22:29<24:13,  2.16s/it, lr: 1.0e-04 loss: 2.364e-01]
anha:  80%|#######9  | 2626/3300 [22:31<24:13,  2.16s/it, lr: 1.0e-04 loss: 6.426e-02]
anha:  80%|#######9  | 2626/3300 [22:31<24:13,  2.16s/it, lr: 1.0e-04 loss: 6.426e-02]
anha:  80%|#######9  | 2627/3300 [22:31<23:35,  2.10s/it, lr: 1.0e-04 loss: 6.426e-02]
anha:  80%|#######9  | 2627/3300 [22:31<23:35,  2.10s/it, lr: 1.0e-04 loss: 6.426e-02]
anha:  80%|#######9  | 2627/3300 [22:32<23:35,  2.10s/it, lr: 1.0e-04 loss: 2.061e-01]
anha:  80%|#######9  | 2627/3300 [22:32<23:35,  2.10s/it, lr: 1.0e-04 loss: 2.061e-01]
anha:  80%|#######9  | 2628/3300 [22:32<20:02,  1.79s/it, lr: 1.0e-04 loss: 2.061e-01]
anha:  80%|#######9  | 2628/3300 [22:32<20:02,  1.79s/it, lr: 1.0e-04 loss: 2.061e-01]
anha:  80%|#######9  | 2628/3300 [22:36<20:02,  1.79s/it, lr: 1.0e-04 loss: 5.077e-02]
anha:  80%|#######9  | 2628/3300 [22:36<20:02,  1.79s/it, lr: 1.0e-04 loss: 5.077e-02]
anha:  80%|#######9  | 2629/3300 [22:36<25:13,  2.26s/it, lr: 1.0e-04 loss: 5.077e-02]
anha:  80%|#######9  | 2629/3300 [22:36<25:13,  2.26s/it, lr: 1.0e-04 loss: 5.077e-02]
anha:  80%|#######9  | 2629/3300 [22:38<25:13,  2.26s/it, lr: 1.0e-04 loss: 4.526e-02]
anha:  80%|#######9  | 2629/3300 [22:38<25:13,  2.26s/it, lr: 1.0e-04 loss: 4.526e-02]
anha:  80%|#######9  | 2630/3300 [22:41<25:11,  2.26s/it, lr: 1.0e-04 loss: 2.584e-01]
anha:  80%|#######9  | 2630/3300 [22:41<25:11,  2.26s/it, lr: 1.0e-04 loss: 2.584e-01]
anha:  80%|#######9  | 2631/3300 [22:41<22:25,  2.01s/it, lr: 1.0e-04 loss: 2.584e-01]
anha:  80%|#######9  | 2631/3300 [22:41<22:25,  2.01s/it, lr: 1.0e-04 loss: 2.584e-01]
anha:  80%|#######9  | 2631/3300 [22:44<22:25,  2.01s/it, lr: 1.0e-04 loss: 1.280e-01]
anha:  80%|#######9  | 2631/3300 [22:44<22:25,  2.01s/it, lr: 1.0e-04 loss: 1.280e-01]
anha:  80%|#######9  | 2632/3300 [22:44<26:07,  2.35s/it, lr: 1.0e-04 loss: 1.280e-01]
anha:  80%|#######9  | 2632/3300 [22:44<26:07,  2.35s/it, lr: 1.0e-04 loss: 1.280e-01]
anha:  80%|#######9  | 2632/3300 [22:46<26:07,  2.35s/it, lr: 1.0e-04 loss: 2.745e-01]
anha:  80%|#######9  | 2632/3300 [22:46<26:07,  2.35s/it, lr: 1.0e-04 loss: 2.745e-01]
anha:  80%|#######9  | 2633/3300 [22:46<25:01,  2.25s/it, lr: 1.0e-04 loss: 2.745e-01]
anha:  80%|#######9  | 2633/3300 [22:46<25:01,  2.25s/it, lr: 1.0e-04 loss: 2.745e-01]
anha:  80%|#######9  | 2633/3300 [22:47<25:01,  2.25s/it, lr: 1.0e-04 loss: 4.967e-02]
anha:  80%|#######9  | 2633/3300 [22:47<25:01,  2.25s/it, lr: 1.0e-04 loss: 4.967e-02]
anha:  80%|#######9  | 2634/3300 [22:47<21:14,  1.91s/it, lr: 1.0e-04 loss: 4.967e-02]
anha:  80%|#######9  | 2634/3300 [22:47<21:14,  1.91s/it, lr: 1.0e-04 loss: 4.967e-02]
anha:  80%|#######9  | 2634/3300 [22:51<21:14,  1.91s/it, lr: 1.0e-04 loss: 5.347e-02]
anha:  80%|#######9  | 2634/3300 [22:51<21:14,  1.91s/it, lr: 1.0e-04 loss: 5.347e-02]
anha:  80%|#######9  | 2635/3300 [22:51<25:43,  2.32s/it, lr: 1.0e-04 loss: 5.347e-02]
anha:  80%|#######9  | 2635/3300 [22:51<25:43,  2.32s/it, lr: 1.0e-04 loss: 5.347e-02]
anha:  80%|#######9  | 2635/3300 [22:52<25:43,  2.32s/it, lr: 1.0e-04 loss: 1.525e-01]
anha:  80%|#######9  | 2635/3300 [22:52<25:43,  2.32s/it, lr: 1.0e-04 loss: 1.525e-01]
anha:  80%|#######9  | 2636/3300 [22:52<21:29,  1.94s/it, lr: 1.0e-04 loss: 1.525e-01]
anha:  80%|#######9  | 2636/3300 [22:52<21:29,  1.94s/it, lr: 1.0e-04 loss: 1.525e-01]
anha:  80%|#######9  | 2636/3300 [22:55<21:29,  1.94s/it, lr: 1.0e-04 loss: 1.132e-01]
anha:  80%|#######9  | 2636/3300 [22:55<21:29,  1.94s/it, lr: 1.0e-04 loss: 1.132e-01]
anha:  80%|#######9  | 2637/3300 [22:55<26:15,  2.38s/it, lr: 1.0e-04 loss: 1.132e-01]
anha:  80%|#######9  | 2637/3300 [22:55<26:15,  2.38s/it, lr: 1.0e-04 loss: 1.132e-01]
anha:  80%|#######9  | 2637/3300 [22:57<26:15,  2.38s/it, lr: 1.0e-04 loss: 3.932e-02]
anha:  80%|#######9  | 2637/3300 [22:57<26:15,  2.38s/it, lr: 1.0e-04 loss: 3.932e-02]
anha:  80%|#######9  | 2638/3300 [22:57<24:56,  2.26s/it, lr: 1.0e-04 loss: 3.932e-02]
anha:  80%|#######9  | 2638/3300 [22:57<24:56,  2.26s/it, lr: 1.0e-04 loss: 3.932e-02]
anha:  80%|#######9  | 2638/3300 [22:58<24:56,  2.26s/it, lr: 1.0e-04 loss: 4.428e-02]
anha:  80%|#######9  | 2638/3300 [22:58<24:56,  2.26s/it, lr: 1.0e-04 loss: 4.428e-02]
anha:  80%|#######9  | 2639/3300 [22:58<21:13,  1.93s/it, lr: 1.0e-04 loss: 4.428e-02]
anha:  80%|#######9  | 2639/3300 [22:58<21:13,  1.93s/it, lr: 1.0e-04 loss: 4.428e-02]
anha:  80%|#######9  | 2639/3300 [22:59<21:13,  1.93s/it, lr: 1.0e-04 loss: 9.136e-02]
anha:  80%|#######9  | 2639/3300 [22:59<21:13,  1.93s/it, lr: 1.0e-04 loss: 9.136e-02]
Saving at step 2640
Saved checkpoint to /workspace/ai-toolkit/output/anha/anha_000002640.safetensors
Saved optimizer to /workspace/ai-toolkit/output/anha/optimizer.pt

anha:  80%|########  | 2640/3300 [23:03<21:11,  1.93s/it, lr: 1.0e-04 loss: 1.619e-01]
anha:  80%|########  | 2640/3300 [23:03<21:11,  1.93s/it, lr: 1.0e-04 loss: 1.619e-01]
anha:  80%|########  | 2641/3300 [23:03<20:32,  1.87s/it, lr: 1.0e-04 loss: 1.619e-01]
anha:  80%|########  | 2641/3300 [23:03<20:32,  1.87s/it, lr: 1.0e-04 loss: 1.619e-01]
anha:  80%|########  | 2641/3300 [23:06<20:32,  1.87s/it, lr: 1.0e-04 loss: 4.480e-02]
anha:  80%|########  | 2641/3300 [23:06<20:32,  1.87s/it, lr: 1.0e-04 loss: 4.480e-02]
anha:  80%|########  | 2642/3300 [23:06<23:59,  2.19s/it, lr: 1.0e-04 loss: 4.480e-02]
anha:  80%|########  | 2642/3300 [23:06<23:59,  2.19s/it, lr: 1.0e-04 loss: 4.480e-02]
anha:  80%|########  | 2642/3300 [23:10<23:59,  2.19s/it, lr: 1.0e-04 loss: 1.585e-01]
anha:  80%|########  | 2642/3300 [23:10<23:59,  2.19s/it, lr: 1.0e-04 loss: 1.585e-01]
anha:  80%|########  | 2643/3300 [23:10<27:38,  2.52s/it, lr: 1.0e-04 loss: 1.585e-01]
anha:  80%|########  | 2643/3300 [23:10<27:38,  2.52s/it, lr: 1.0e-04 loss: 1.585e-01]
anha:  80%|########  | 2643/3300 [23:12<27:38,  2.52s/it, lr: 1.0e-04 loss: 5.795e-02]
anha:  80%|########  | 2643/3300 [23:12<27:38,  2.52s/it, lr: 1.0e-04 loss: 5.795e-02]
anha:  80%|########  | 2644/3300 [23:12<25:58,  2.38s/it, lr: 1.0e-04 loss: 5.795e-02]
anha:  80%|########  | 2644/3300 [23:12<25:58,  2.38s/it, lr: 1.0e-04 loss: 5.795e-02]
anha:  80%|########  | 2644/3300 [23:13<25:58,  2.38s/it, lr: 1.0e-04 loss: 3.000e-02]
anha:  80%|########  | 2644/3300 [23:13<25:58,  2.38s/it, lr: 1.0e-04 loss: 3.000e-02]
anha:  80%|########  | 2645/3300 [23:13<21:48,  2.00s/it, lr: 1.0e-04 loss: 3.000e-02]
anha:  80%|########  | 2645/3300 [23:13<21:48,  2.00s/it, lr: 1.0e-04 loss: 3.000e-02]
anha:  80%|########  | 2645/3300 [23:15<21:48,  2.00s/it, lr: 1.0e-04 loss: 5.219e-02]
anha:  80%|########  | 2645/3300 [23:15<21:48,  2.00s/it, lr: 1.0e-04 loss: 5.219e-02]
anha:  80%|########  | 2646/3300 [23:15<21:43,  1.99s/it, lr: 1.0e-04 loss: 5.219e-02]
anha:  80%|########  | 2646/3300 [23:15<21:43,  1.99s/it, lr: 1.0e-04 loss: 5.219e-02]
anha:  80%|########  | 2646/3300 [23:16<21:43,  1.99s/it, lr: 1.0e-04 loss: 5.482e-02]
anha:  80%|########  | 2646/3300 [23:16<21:43,  1.99s/it, lr: 1.0e-04 loss: 5.482e-02]
anha:  80%|########  | 2647/3300 [23:16<20:48,  1.91s/it, lr: 1.0e-04 loss: 5.482e-02]
anha:  80%|########  | 2647/3300 [23:16<20:48,  1.91s/it, lr: 1.0e-04 loss: 5.482e-02]
anha:  80%|########  | 2647/3300 [23:20<20:48,  1.91s/it, lr: 1.0e-04 loss: 9.137e-02]
anha:  80%|########  | 2647/3300 [23:20<20:48,  1.91s/it, lr: 1.0e-04 loss: 9.137e-02]
anha:  80%|########  | 2648/3300 [23:20<25:45,  2.37s/it, lr: 1.0e-04 loss: 9.137e-02]
anha:  80%|########  | 2648/3300 [23:20<25:45,  2.37s/it, lr: 1.0e-04 loss: 9.137e-02]
anha:  80%|########  | 2648/3300 [23:21<25:45,  2.37s/it, lr: 1.0e-04 loss: 2.933e-02]
anha:  80%|########  | 2648/3300 [23:21<25:45,  2.37s/it, lr: 1.0e-04 loss: 2.933e-02]
anha:  80%|########  | 2649/3300 [23:21<21:20,  1.97s/it, lr: 1.0e-04 loss: 2.933e-02]
anha:  80%|########  | 2649/3300 [23:21<21:20,  1.97s/it, lr: 1.0e-04 loss: 2.933e-02]
anha:  80%|########  | 2649/3300 [23:24<21:20,  1.97s/it, lr: 1.0e-04 loss: 2.167e-01]
anha:  80%|########  | 2649/3300 [23:24<21:20,  1.97s/it, lr: 1.0e-04 loss: 2.167e-01]
anha:  80%|########  | 2650/3300 [23:25<21:18,  1.97s/it, lr: 1.0e-04 loss: 4.883e-02]
anha:  80%|########  | 2650/3300 [23:25<21:18,  1.97s/it, lr: 1.0e-04 loss: 4.883e-02]
anha:  80%|########  | 2651/3300 [23:25<13:59,  1.29s/it, lr: 1.0e-04 loss: 4.883e-02]
anha:  80%|########  | 2651/3300 [23:25<13:59,  1.29s/it, lr: 1.0e-04 loss: 4.883e-02]
anha:  80%|########  | 2651/3300 [23:29<13:59,  1.29s/it, lr: 1.0e-04 loss: 3.119e-01]
anha:  80%|########  | 2651/3300 [23:29<13:59,  1.29s/it, lr: 1.0e-04 loss: 3.119e-01]
anha:  80%|########  | 2652/3300 [23:29<19:29,  1.81s/it, lr: 1.0e-04 loss: 3.119e-01]
anha:  80%|########  | 2652/3300 [23:29<19:29,  1.81s/it, lr: 1.0e-04 loss: 3.119e-01]
anha:  80%|########  | 2652/3300 [23:30<19:29,  1.81s/it, lr: 1.0e-04 loss: 4.518e-02]
anha:  80%|########  | 2652/3300 [23:30<19:29,  1.81s/it, lr: 1.0e-04 loss: 4.518e-02]
anha:  80%|########  | 2653/3300 [23:30<17:14,  1.60s/it, lr: 1.0e-04 loss: 4.518e-02]
anha:  80%|########  | 2653/3300 [23:30<17:14,  1.60s/it, lr: 1.0e-04 loss: 4.518e-02]
anha:  80%|########  | 2653/3300 [23:33<17:14,  1.60s/it, lr: 1.0e-04 loss: 7.871e-02]
anha:  80%|########  | 2653/3300 [23:33<17:14,  1.60s/it, lr: 1.0e-04 loss: 7.871e-02]
anha:  80%|########  | 2654/3300 [23:33<22:23,  2.08s/it, lr: 1.0e-04 loss: 7.871e-02]
anha:  80%|########  | 2654/3300 [23:33<22:23,  2.08s/it, lr: 1.0e-04 loss: 7.871e-02]
anha:  80%|########  | 2654/3300 [23:36<22:23,  2.08s/it, lr: 1.0e-04 loss: 9.540e-02]
anha:  80%|########  | 2654/3300 [23:36<22:23,  2.08s/it, lr: 1.0e-04 loss: 9.540e-02]
anha:  80%|########  | 2655/3300 [23:36<25:49,  2.40s/it, lr: 1.0e-04 loss: 9.540e-02]
anha:  80%|########  | 2655/3300 [23:36<25:49,  2.40s/it, lr: 1.0e-04 loss: 9.540e-02]
anha:  80%|########  | 2655/3300 [23:38<25:49,  2.40s/it, lr: 1.0e-04 loss: 1.244e-01]
anha:  80%|########  | 2655/3300 [23:38<25:49,  2.40s/it, lr: 1.0e-04 loss: 1.244e-01]
anha:  80%|########  | 2656/3300 [23:38<23:44,  2.21s/it, lr: 1.0e-04 loss: 1.244e-01]
anha:  80%|########  | 2656/3300 [23:38<23:44,  2.21s/it, lr: 1.0e-04 loss: 1.244e-01]
anha:  80%|########  | 2656/3300 [23:41<23:44,  2.21s/it, lr: 1.0e-04 loss: 6.954e-02]
anha:  80%|########  | 2656/3300 [23:41<23:44,  2.21s/it, lr: 1.0e-04 loss: 6.954e-02]
anha:  81%|########  | 2657/3300 [23:41<27:02,  2.52s/it, lr: 1.0e-04 loss: 6.954e-02]
anha:  81%|########  | 2657/3300 [23:41<27:02,  2.52s/it, lr: 1.0e-04 loss: 6.954e-02]
anha:  81%|########  | 2657/3300 [23:43<27:02,  2.52s/it, lr: 1.0e-04 loss: 1.735e-01]
anha:  81%|########  | 2657/3300 [23:43<27:02,  2.52s/it, lr: 1.0e-04 loss: 1.735e-01]
anha:  81%|########  | 2658/3300 [23:43<25:17,  2.36s/it, lr: 1.0e-04 loss: 1.735e-01]
anha:  81%|########  | 2658/3300 [23:43<25:17,  2.36s/it, lr: 1.0e-04 loss: 1.735e-01]
anha:  81%|########  | 2658/3300 [23:45<25:17,  2.36s/it, lr: 1.0e-04 loss: 5.164e-02]
anha:  81%|########  | 2658/3300 [23:45<25:17,  2.36s/it, lr: 1.0e-04 loss: 5.164e-02]
anha:  81%|########  | 2659/3300 [23:45<24:16,  2.27s/it, lr: 1.0e-04 loss: 5.164e-02]
anha:  81%|########  | 2659/3300 [23:45<24:16,  2.27s/it, lr: 1.0e-04 loss: 5.164e-02]
anha:  81%|########  | 2659/3300 [23:46<24:16,  2.27s/it, lr: 1.0e-04 loss: 6.279e-02]
anha:  81%|########  | 2659/3300 [23:46<24:16,  2.27s/it, lr: 1.0e-04 loss: 6.279e-02]
anha:  81%|########  | 2660/3300 [23:48<24:14,  2.27s/it, lr: 1.0e-04 loss: 2.701e-02]
anha:  81%|########  | 2660/3300 [23:48<24:14,  2.27s/it, lr: 1.0e-04 loss: 2.701e-02]
anha:  81%|########  | 2661/3300 [23:48<17:53,  1.68s/it, lr: 1.0e-04 loss: 2.701e-02]
anha:  81%|########  | 2661/3300 [23:48<17:53,  1.68s/it, lr: 1.0e-04 loss: 2.701e-02]
anha:  81%|########  | 2661/3300 [23:49<17:53,  1.68s/it, lr: 1.0e-04 loss: 4.500e-02]
anha:  81%|########  | 2661/3300 [23:49<17:53,  1.68s/it, lr: 1.0e-04 loss: 4.500e-02]
anha:  81%|########  | 2662/3300 [23:49<16:04,  1.51s/it, lr: 1.0e-04 loss: 4.500e-02]
anha:  81%|########  | 2662/3300 [23:49<16:04,  1.51s/it, lr: 1.0e-04 loss: 4.500e-02]
anha:  81%|########  | 2662/3300 [23:52<16:04,  1.51s/it, lr: 1.0e-04 loss: 1.513e-01]
anha:  81%|########  | 2662/3300 [23:52<16:04,  1.51s/it, lr: 1.0e-04 loss: 1.513e-01]
anha:  81%|########  | 2663/3300 [23:52<20:56,  1.97s/it, lr: 1.0e-04 loss: 1.513e-01]
anha:  81%|########  | 2663/3300 [23:52<20:56,  1.97s/it, lr: 1.0e-04 loss: 1.513e-01]
anha:  81%|########  | 2663/3300 [23:54<20:56,  1.97s/it, lr: 1.0e-04 loss: 4.847e-02]
anha:  81%|########  | 2663/3300 [23:54<20:56,  1.97s/it, lr: 1.0e-04 loss: 4.847e-02]
anha:  81%|########  | 2664/3300 [23:54<20:59,  1.98s/it, lr: 1.0e-04 loss: 4.847e-02]
anha:  81%|########  | 2664/3300 [23:54<20:59,  1.98s/it, lr: 1.0e-04 loss: 4.847e-02]
anha:  81%|########  | 2664/3300 [23:55<20:59,  1.98s/it, lr: 1.0e-04 loss: 8.326e-02]
anha:  81%|########  | 2664/3300 [23:55<20:59,  1.98s/it, lr: 1.0e-04 loss: 8.326e-02]
anha:  81%|########  | 2665/3300 [23:55<18:05,  1.71s/it, lr: 1.0e-04 loss: 8.326e-02]
anha:  81%|########  | 2665/3300 [23:55<18:05,  1.71s/it, lr: 1.0e-04 loss: 8.326e-02]
anha:  81%|########  | 2665/3300 [23:59<18:05,  1.71s/it, lr: 1.0e-04 loss: 6.292e-02]
anha:  81%|########  | 2665/3300 [23:59<18:05,  1.71s/it, lr: 1.0e-04 loss: 6.292e-02]
anha:  81%|########  | 2666/3300 [23:59<23:16,  2.20s/it, lr: 1.0e-04 loss: 6.292e-02]
anha:  81%|########  | 2666/3300 [23:59<23:16,  2.20s/it, lr: 1.0e-04 loss: 6.292e-02]
anha:  81%|########  | 2666/3300 [24:01<23:16,  2.20s/it, lr: 1.0e-04 loss: 1.063e-01]
anha:  81%|########  | 2666/3300 [24:01<23:16,  2.20s/it, lr: 1.0e-04 loss: 1.063e-01]
anha:  81%|########  | 2667/3300 [24:01<21:49,  2.07s/it, lr: 1.0e-04 loss: 1.063e-01]
anha:  81%|########  | 2667/3300 [24:01<21:49,  2.07s/it, lr: 1.0e-04 loss: 1.063e-01]
anha:  81%|########  | 2667/3300 [24:02<21:49,  2.07s/it, lr: 1.0e-04 loss: 6.768e-02]
anha:  81%|########  | 2667/3300 [24:02<21:49,  2.07s/it, lr: 1.0e-04 loss: 6.768e-02]
anha:  81%|########  | 2668/3300 [24:02<18:31,  1.76s/it, lr: 1.0e-04 loss: 6.768e-02]
anha:  81%|########  | 2668/3300 [24:02<18:31,  1.76s/it, lr: 1.0e-04 loss: 6.768e-02]
anha:  81%|########  | 2668/3300 [24:04<18:31,  1.76s/it, lr: 1.0e-04 loss: 4.282e-02]
anha:  81%|########  | 2668/3300 [24:04<18:31,  1.76s/it, lr: 1.0e-04 loss: 4.282e-02]
anha:  81%|########  | 2669/3300 [24:04<19:18,  1.84s/it, lr: 1.0e-04 loss: 4.282e-02]
anha:  81%|########  | 2669/3300 [24:04<19:18,  1.84s/it, lr: 1.0e-04 loss: 4.282e-02]
anha:  81%|########  | 2669/3300 [24:05<19:18,  1.84s/it, lr: 1.0e-04 loss: 6.916e-02]
anha:  81%|########  | 2669/3300 [24:05<19:18,  1.84s/it, lr: 1.0e-04 loss: 6.916e-02]
anha:  81%|########  | 2670/3300 [24:06<19:16,  1.84s/it, lr: 1.0e-04 loss: 6.387e-02]
anha:  81%|########  | 2670/3300 [24:06<19:16,  1.84s/it, lr: 1.0e-04 loss: 6.387e-02]
anha:  81%|########  | 2671/3300 [24:06<12:48,  1.22s/it, lr: 1.0e-04 loss: 6.387e-02]
anha:  81%|########  | 2671/3300 [24:06<12:48,  1.22s/it, lr: 1.0e-04 loss: 6.387e-02]
anha:  81%|########  | 2671/3300 [24:08<12:48,  1.22s/it, lr: 1.0e-04 loss: 3.590e-02]
anha:  81%|########  | 2671/3300 [24:08<12:48,  1.22s/it, lr: 1.0e-04 loss: 3.590e-02]
anha:  81%|########  | 2672/3300 [24:08<14:40,  1.40s/it, lr: 1.0e-04 loss: 3.590e-02]
anha:  81%|########  | 2672/3300 [24:08<14:40,  1.40s/it, lr: 1.0e-04 loss: 3.590e-02]
anha:  81%|########  | 2672/3300 [24:09<14:40,  1.40s/it, lr: 1.0e-04 loss: 4.147e-02]
anha:  81%|########  | 2672/3300 [24:09<14:40,  1.40s/it, lr: 1.0e-04 loss: 4.147e-02]
anha:  81%|########1 | 2673/3300 [24:09<13:34,  1.30s/it, lr: 1.0e-04 loss: 4.147e-02]
anha:  81%|########1 | 2673/3300 [24:09<13:34,  1.30s/it, lr: 1.0e-04 loss: 4.147e-02]
anha:  81%|########1 | 2673/3300 [24:10<13:34,  1.30s/it, lr: 1.0e-04 loss: 6.316e-02]
anha:  81%|########1 | 2673/3300 [24:10<13:34,  1.30s/it, lr: 1.0e-04 loss: 6.316e-02]
anha:  81%|########1 | 2674/3300 [24:10<12:44,  1.22s/it, lr: 1.0e-04 loss: 6.316e-02]
anha:  81%|########1 | 2674/3300 [24:10<12:44,  1.22s/it, lr: 1.0e-04 loss: 6.316e-02]
anha:  81%|########1 | 2674/3300 [24:11<12:44,  1.22s/it, lr: 1.0e-04 loss: 3.410e-02]
anha:  81%|########1 | 2674/3300 [24:11<12:44,  1.22s/it, lr: 1.0e-04 loss: 3.410e-02]
anha:  81%|########1 | 2675/3300 [24:11<14:10,  1.36s/it, lr: 1.0e-04 loss: 3.410e-02]
anha:  81%|########1 | 2675/3300 [24:11<14:10,  1.36s/it, lr: 1.0e-04 loss: 3.410e-02]
anha:  81%|########1 | 2675/3300 [24:12<14:10,  1.36s/it, lr: 1.0e-04 loss: 1.254e-01]
anha:  81%|########1 | 2675/3300 [24:12<14:10,  1.36s/it, lr: 1.0e-04 loss: 1.254e-01]
anha:  81%|########1 | 2676/3300 [24:12<13:13,  1.27s/it, lr: 1.0e-04 loss: 1.254e-01]
anha:  81%|########1 | 2676/3300 [24:12<13:13,  1.27s/it, lr: 1.0e-04 loss: 1.254e-01]
anha:  81%|########1 | 2676/3300 [24:16<13:13,  1.27s/it, lr: 1.0e-04 loss: 3.219e-02]
anha:  81%|########1 | 2676/3300 [24:16<13:13,  1.27s/it, lr: 1.0e-04 loss: 3.219e-02]
anha:  81%|########1 | 2677/3300 [24:16<19:30,  1.88s/it, lr: 1.0e-04 loss: 3.219e-02]
anha:  81%|########1 | 2677/3300 [24:16<19:30,  1.88s/it, lr: 1.0e-04 loss: 3.219e-02]
anha:  81%|########1 | 2677/3300 [24:18<19:30,  1.88s/it, lr: 1.0e-04 loss: 5.504e-02]
anha:  81%|########1 | 2677/3300 [24:18<19:30,  1.88s/it, lr: 1.0e-04 loss: 5.504e-02]
anha:  81%|########1 | 2678/3300 [24:18<19:58,  1.93s/it, lr: 1.0e-04 loss: 5.504e-02]
anha:  81%|########1 | 2678/3300 [24:18<19:58,  1.93s/it, lr: 1.0e-04 loss: 5.504e-02]
anha:  81%|########1 | 2678/3300 [24:21<19:58,  1.93s/it, lr: 1.0e-04 loss: 5.904e-02]
anha:  81%|########1 | 2678/3300 [24:21<19:58,  1.93s/it, lr: 1.0e-04 loss: 5.904e-02]
anha:  81%|########1 | 2679/3300 [24:21<24:05,  2.33s/it, lr: 1.0e-04 loss: 5.904e-02]
anha:  81%|########1 | 2679/3300 [24:21<24:05,  2.33s/it, lr: 1.0e-04 loss: 5.904e-02]
anha:  81%|########1 | 2679/3300 [24:23<24:05,  2.33s/it, lr: 1.0e-04 loss: 8.436e-02]
anha:  81%|########1 | 2679/3300 [24:23<24:05,  2.33s/it, lr: 1.0e-04 loss: 8.436e-02]
anha:  81%|########1 | 2680/3300 [24:24<24:02,  2.33s/it, lr: 1.0e-04 loss: 3.512e-02]
anha:  81%|########1 | 2680/3300 [24:24<24:02,  2.33s/it, lr: 1.0e-04 loss: 3.512e-02]
anha:  81%|########1 | 2681/3300 [24:24<15:23,  1.49s/it, lr: 1.0e-04 loss: 3.512e-02]
anha:  81%|########1 | 2681/3300 [24:24<15:23,  1.49s/it, lr: 1.0e-04 loss: 3.512e-02]
anha:  81%|########1 | 2681/3300 [24:25<15:23,  1.49s/it, lr: 1.0e-04 loss: 6.542e-02]
anha:  81%|########1 | 2681/3300 [24:25<15:23,  1.49s/it, lr: 1.0e-04 loss: 6.542e-02]
anha:  81%|########1 | 2682/3300 [24:25<14:06,  1.37s/it, lr: 1.0e-04 loss: 6.542e-02]
anha:  81%|########1 | 2682/3300 [24:25<14:06,  1.37s/it, lr: 1.0e-04 loss: 6.542e-02]
anha:  81%|########1 | 2682/3300 [24:28<14:06,  1.37s/it, lr: 1.0e-04 loss: 7.161e-02]
anha:  81%|########1 | 2682/3300 [24:28<14:06,  1.37s/it, lr: 1.0e-04 loss: 7.161e-02]
anha:  81%|########1 | 2683/3300 [24:28<17:08,  1.67s/it, lr: 1.0e-04 loss: 7.161e-02]
anha:  81%|########1 | 2683/3300 [24:28<17:08,  1.67s/it, lr: 1.0e-04 loss: 7.161e-02]
anha:  81%|########1 | 2683/3300 [24:30<17:08,  1.67s/it, lr: 1.0e-04 loss: 3.986e-02]
anha:  81%|########1 | 2683/3300 [24:30<17:08,  1.67s/it, lr: 1.0e-04 loss: 3.986e-02]
anha:  81%|########1 | 2684/3300 [24:30<17:58,  1.75s/it, lr: 1.0e-04 loss: 3.986e-02]
anha:  81%|########1 | 2684/3300 [24:30<17:58,  1.75s/it, lr: 1.0e-04 loss: 3.986e-02]
anha:  81%|########1 | 2684/3300 [24:31<17:58,  1.75s/it, lr: 1.0e-04 loss: 6.674e-02]
anha:  81%|########1 | 2684/3300 [24:31<17:58,  1.75s/it, lr: 1.0e-04 loss: 6.674e-02]
anha:  81%|########1 | 2685/3300 [24:31<17:57,  1.75s/it, lr: 1.0e-04 loss: 6.674e-02]
anha:  81%|########1 | 2685/3300 [24:31<17:57,  1.75s/it, lr: 1.0e-04 loss: 6.674e-02]
anha:  81%|########1 | 2685/3300 [24:35<17:57,  1.75s/it, lr: 1.0e-04 loss: 5.282e-02]
anha:  81%|########1 | 2685/3300 [24:35<17:57,  1.75s/it, lr: 1.0e-04 loss: 5.282e-02]
anha:  81%|########1 | 2686/3300 [24:35<22:58,  2.24s/it, lr: 1.0e-04 loss: 5.282e-02]
anha:  81%|########1 | 2686/3300 [24:35<22:58,  2.24s/it, lr: 1.0e-04 loss: 5.282e-02]
anha:  81%|########1 | 2686/3300 [24:36<22:58,  2.24s/it, lr: 1.0e-04 loss: 8.825e-02]
anha:  81%|########1 | 2686/3300 [24:36<22:58,  2.24s/it, lr: 1.0e-04 loss: 8.825e-02]
anha:  81%|########1 | 2687/3300 [24:36<19:38,  1.92s/it, lr: 1.0e-04 loss: 8.825e-02]
anha:  81%|########1 | 2687/3300 [24:36<19:38,  1.92s/it, lr: 1.0e-04 loss: 8.825e-02]
anha:  81%|########1 | 2687/3300 [24:37<19:38,  1.92s/it, lr: 1.0e-04 loss: 7.816e-02]
anha:  81%|########1 | 2687/3300 [24:37<19:38,  1.92s/it, lr: 1.0e-04 loss: 7.816e-02]
anha:  81%|########1 | 2688/3300 [24:37<16:53,  1.66s/it, lr: 1.0e-04 loss: 7.816e-02]
anha:  81%|########1 | 2688/3300 [24:37<16:53,  1.66s/it, lr: 1.0e-04 loss: 7.816e-02]
anha:  81%|########1 | 2688/3300 [24:40<16:53,  1.66s/it, lr: 1.0e-04 loss: 3.173e-02]
anha:  81%|########1 | 2688/3300 [24:40<16:53,  1.66s/it, lr: 1.0e-04 loss: 3.173e-02]
anha:  81%|########1 | 2689/3300 [24:40<22:00,  2.16s/it, lr: 1.0e-04 loss: 3.173e-02]
anha:  81%|########1 | 2689/3300 [24:40<22:00,  2.16s/it, lr: 1.0e-04 loss: 3.173e-02]
anha:  81%|########1 | 2689/3300 [24:42<22:00,  2.16s/it, lr: 1.0e-04 loss: 1.217e-01]
anha:  81%|########1 | 2689/3300 [24:42<22:00,  2.16s/it, lr: 1.0e-04 loss: 1.217e-01]
anha:  82%|########1 | 2690/3300 [24:43<21:58,  2.16s/it, lr: 1.0e-04 loss: 6.206e-02]
anha:  82%|########1 | 2690/3300 [24:43<21:58,  2.16s/it, lr: 1.0e-04 loss: 6.206e-02]
anha:  82%|########1 | 2691/3300 [24:43<14:14,  1.40s/it, lr: 1.0e-04 loss: 6.206e-02]
anha:  82%|########1 | 2691/3300 [24:43<14:14,  1.40s/it, lr: 1.0e-04 loss: 6.206e-02]
anha:  82%|########1 | 2691/3300 [24:46<14:14,  1.40s/it, lr: 1.0e-04 loss: 3.705e-02]
anha:  82%|########1 | 2691/3300 [24:46<14:14,  1.40s/it, lr: 1.0e-04 loss: 3.705e-02]
anha:  82%|########1 | 2692/3300 [24:46<19:07,  1.89s/it, lr: 1.0e-04 loss: 3.705e-02]
anha:  82%|########1 | 2692/3300 [24:46<19:07,  1.89s/it, lr: 1.0e-04 loss: 3.705e-02]
anha:  82%|########1 | 2692/3300 [24:48<19:07,  1.89s/it, lr: 1.0e-04 loss: 4.058e-02]
anha:  82%|########1 | 2692/3300 [24:48<19:07,  1.89s/it, lr: 1.0e-04 loss: 4.058e-02]
anha:  82%|########1 | 2693/3300 [24:48<19:20,  1.91s/it, lr: 1.0e-04 loss: 4.058e-02]
anha:  82%|########1 | 2693/3300 [24:48<19:20,  1.91s/it, lr: 1.0e-04 loss: 4.058e-02]
anha:  82%|########1 | 2693/3300 [24:50<19:20,  1.91s/it, lr: 1.0e-04 loss: 4.736e-02]
anha:  82%|########1 | 2693/3300 [24:50<19:20,  1.91s/it, lr: 1.0e-04 loss: 4.736e-02]
anha:  82%|########1 | 2694/3300 [24:50<19:25,  1.92s/it, lr: 1.0e-04 loss: 4.736e-02]
anha:  82%|########1 | 2694/3300 [24:50<19:25,  1.92s/it, lr: 1.0e-04 loss: 4.736e-02]
anha:  82%|########1 | 2694/3300 [24:54<19:25,  1.92s/it, lr: 1.0e-04 loss: 4.601e-02]
anha:  82%|########1 | 2694/3300 [24:54<19:25,  1.92s/it, lr: 1.0e-04 loss: 4.601e-02]
anha:  82%|########1 | 2695/3300 [24:54<23:45,  2.36s/it, lr: 1.0e-04 loss: 4.601e-02]
anha:  82%|########1 | 2695/3300 [24:54<23:45,  2.36s/it, lr: 1.0e-04 loss: 4.601e-02]
anha:  82%|########1 | 2695/3300 [24:57<23:45,  2.36s/it, lr: 1.0e-04 loss: 1.185e-01]
anha:  82%|########1 | 2695/3300 [24:57<23:45,  2.36s/it, lr: 1.0e-04 loss: 1.185e-01]
anha:  82%|########1 | 2696/3300 [24:57<26:59,  2.68s/it, lr: 1.0e-04 loss: 1.185e-01]
anha:  82%|########1 | 2696/3300 [24:57<26:59,  2.68s/it, lr: 1.0e-04 loss: 1.185e-01]
anha:  82%|########1 | 2696/3300 [24:58<26:59,  2.68s/it, lr: 1.0e-04 loss: 5.987e-02]
anha:  82%|########1 | 2696/3300 [24:58<26:59,  2.68s/it, lr: 1.0e-04 loss: 5.987e-02]
anha:  82%|########1 | 2697/3300 [24:58<22:27,  2.24s/it, lr: 1.0e-04 loss: 5.987e-02]
anha:  82%|########1 | 2697/3300 [24:58<22:27,  2.24s/it, lr: 1.0e-04 loss: 5.987e-02]
anha:  82%|########1 | 2697/3300 [25:02<22:27,  2.24s/it, lr: 1.0e-04 loss: 7.242e-02]
anha:  82%|########1 | 2697/3300 [25:02<22:27,  2.24s/it, lr: 1.0e-04 loss: 7.242e-02]
anha:  82%|########1 | 2698/3300 [25:02<25:21,  2.53s/it, lr: 1.0e-04 loss: 7.242e-02]
anha:  82%|########1 | 2698/3300 [25:02<25:21,  2.53s/it, lr: 1.0e-04 loss: 7.242e-02]
anha:  82%|########1 | 2698/3300 [25:03<25:21,  2.53s/it, lr: 1.0e-04 loss: 1.040e-01]
anha:  82%|########1 | 2698/3300 [25:03<25:21,  2.53s/it, lr: 1.0e-04 loss: 1.040e-01]
anha:  82%|########1 | 2699/3300 [25:03<20:47,  2.08s/it, lr: 1.0e-04 loss: 1.040e-01]
anha:  82%|########1 | 2699/3300 [25:03<20:47,  2.08s/it, lr: 1.0e-04 loss: 1.040e-01]
anha:  82%|########1 | 2699/3300 [25:04<20:47,  2.08s/it, lr: 1.0e-04 loss: 5.664e-02]
anha:  82%|########1 | 2699/3300 [25:04<20:47,  2.08s/it, lr: 1.0e-04 loss: 5.664e-02]
anha:  82%|########1 | 2700/3300 [25:07<20:45,  2.08s/it, lr: 1.0e-04 loss: 5.667e-02]
anha:  82%|########1 | 2700/3300 [25:07<20:45,  2.08s/it, lr: 1.0e-04 loss: 5.667e-02]
anha:  82%|########1 | 2701/3300 [25:07<18:56,  1.90s/it, lr: 1.0e-04 loss: 5.667e-02]
anha:  82%|########1 | 2701/3300 [25:07<18:56,  1.90s/it, lr: 1.0e-04 loss: 5.667e-02]
anha:  82%|########1 | 2701/3300 [25:09<18:56,  1.90s/it, lr: 1.0e-04 loss: 3.415e-01]
anha:  82%|########1 | 2701/3300 [25:09<18:56,  1.90s/it, lr: 1.0e-04 loss: 3.415e-01]
anha:  82%|########1 | 2702/3300 [25:09<19:02,  1.91s/it, lr: 1.0e-04 loss: 3.415e-01]
anha:  82%|########1 | 2702/3300 [25:09<19:02,  1.91s/it, lr: 1.0e-04 loss: 3.415e-01]
anha:  82%|########1 | 2702/3300 [25:12<19:02,  1.91s/it, lr: 1.0e-04 loss: 6.445e-02]
anha:  82%|########1 | 2702/3300 [25:12<19:02,  1.91s/it, lr: 1.0e-04 loss: 6.445e-02]
anha:  82%|########1 | 2703/3300 [25:12<22:47,  2.29s/it, lr: 1.0e-04 loss: 6.445e-02]
anha:  82%|########1 | 2703/3300 [25:12<22:47,  2.29s/it, lr: 1.0e-04 loss: 6.445e-02]
anha:  82%|########1 | 2703/3300 [25:14<22:47,  2.29s/it, lr: 1.0e-04 loss: 3.687e-02]
anha:  82%|########1 | 2703/3300 [25:14<22:47,  2.29s/it, lr: 1.0e-04 loss: 3.687e-02]
anha:  82%|########1 | 2704/3300 [25:14<21:54,  2.20s/it, lr: 1.0e-04 loss: 3.687e-02]
anha:  82%|########1 | 2704/3300 [25:14<21:54,  2.20s/it, lr: 1.0e-04 loss: 3.687e-02]
anha:  82%|########1 | 2704/3300 [25:16<21:54,  2.20s/it, lr: 1.0e-04 loss: 5.046e-02]
anha:  82%|########1 | 2704/3300 [25:16<21:54,  2.20s/it, lr: 1.0e-04 loss: 5.046e-02]
anha:  82%|########1 | 2705/3300 [25:16<20:35,  2.08s/it, lr: 1.0e-04 loss: 5.046e-02]
anha:  82%|########1 | 2705/3300 [25:16<20:35,  2.08s/it, lr: 1.0e-04 loss: 5.046e-02]
anha:  82%|########1 | 2705/3300 [25:18<20:35,  2.08s/it, lr: 1.0e-04 loss: 5.925e-02]
anha:  82%|########1 | 2705/3300 [25:18<20:35,  2.08s/it, lr: 1.0e-04 loss: 5.925e-02]
anha:  82%|########2 | 2706/3300 [25:18<20:32,  2.08s/it, lr: 1.0e-04 loss: 5.925e-02]
anha:  82%|########2 | 2706/3300 [25:18<20:32,  2.08s/it, lr: 1.0e-04 loss: 5.925e-02]
anha:  82%|########2 | 2706/3300 [25:19<20:32,  2.08s/it, lr: 1.0e-04 loss: 4.112e-02]
anha:  82%|########2 | 2706/3300 [25:19<20:32,  2.08s/it, lr: 1.0e-04 loss: 4.112e-02]
anha:  82%|########2 | 2707/3300 [25:19<17:25,  1.76s/it, lr: 1.0e-04 loss: 4.112e-02]
anha:  82%|########2 | 2707/3300 [25:19<17:25,  1.76s/it, lr: 1.0e-04 loss: 4.112e-02]
anha:  82%|########2 | 2707/3300 [25:23<17:25,  1.76s/it, lr: 1.0e-04 loss: 1.938e-01]
anha:  82%|########2 | 2707/3300 [25:23<17:25,  1.76s/it, lr: 1.0e-04 loss: 1.938e-01]
anha:  82%|########2 | 2708/3300 [25:23<21:47,  2.21s/it, lr: 1.0e-04 loss: 1.938e-01]
anha:  82%|########2 | 2708/3300 [25:23<21:47,  2.21s/it, lr: 1.0e-04 loss: 1.938e-01]
anha:  82%|########2 | 2708/3300 [25:24<21:47,  2.21s/it, lr: 1.0e-04 loss: 1.549e-01]
anha:  82%|########2 | 2708/3300 [25:24<21:47,  2.21s/it, lr: 1.0e-04 loss: 1.549e-01]
anha:  82%|########2 | 2709/3300 [25:24<20:26,  2.08s/it, lr: 1.0e-04 loss: 1.549e-01]
anha:  82%|########2 | 2709/3300 [25:24<20:26,  2.08s/it, lr: 1.0e-04 loss: 1.549e-01]
anha:  82%|########2 | 2709/3300 [25:28<20:26,  2.08s/it, lr: 1.0e-04 loss: 1.269e-01]
anha:  82%|########2 | 2709/3300 [25:28<20:26,  2.08s/it, lr: 1.0e-04 loss: 1.269e-01]
anha:  82%|########2 | 2710/3300 [25:29<20:24,  2.08s/it, lr: 1.0e-04 loss: 5.027e-02]
anha:  82%|########2 | 2710/3300 [25:29<20:24,  2.08s/it, lr: 1.0e-04 loss: 5.027e-02]
anha:  82%|########2 | 2711/3300 [25:29<13:18,  1.36s/it, lr: 1.0e-04 loss: 5.027e-02]
anha:  82%|########2 | 2711/3300 [25:29<13:18,  1.36s/it, lr: 1.0e-04 loss: 5.027e-02]
anha:  82%|########2 | 2711/3300 [25:32<13:18,  1.36s/it, lr: 1.0e-04 loss: 6.934e-02]
anha:  82%|########2 | 2711/3300 [25:32<13:18,  1.36s/it, lr: 1.0e-04 loss: 6.934e-02]
anha:  82%|########2 | 2712/3300 [25:32<18:24,  1.88s/it, lr: 1.0e-04 loss: 6.934e-02]
anha:  82%|########2 | 2712/3300 [25:32<18:24,  1.88s/it, lr: 1.0e-04 loss: 6.934e-02]
anha:  82%|########2 | 2712/3300 [25:33<18:24,  1.88s/it, lr: 1.0e-04 loss: 5.860e-02]
anha:  82%|########2 | 2712/3300 [25:33<18:24,  1.88s/it, lr: 1.0e-04 loss: 5.860e-02]
anha:  82%|########2 | 2713/3300 [25:33<16:10,  1.65s/it, lr: 1.0e-04 loss: 5.860e-02]
anha:  82%|########2 | 2713/3300 [25:33<16:10,  1.65s/it, lr: 1.0e-04 loss: 5.860e-02]
anha:  82%|########2 | 2713/3300 [25:36<16:10,  1.65s/it, lr: 1.0e-04 loss: 6.547e-02]
anha:  82%|########2 | 2713/3300 [25:36<16:10,  1.65s/it, lr: 1.0e-04 loss: 6.547e-02]
anha:  82%|########2 | 2714/3300 [25:36<20:07,  2.06s/it, lr: 1.0e-04 loss: 6.547e-02]
anha:  82%|########2 | 2714/3300 [25:36<20:07,  2.06s/it, lr: 1.0e-04 loss: 6.547e-02]
anha:  82%|########2 | 2714/3300 [25:38<20:07,  2.06s/it, lr: 1.0e-04 loss: 6.243e-02]
anha:  82%|########2 | 2714/3300 [25:38<20:07,  2.06s/it, lr: 1.0e-04 loss: 6.243e-02]
anha:  82%|########2 | 2715/3300 [25:38<17:38,  1.81s/it, lr: 1.0e-04 loss: 6.243e-02]
anha:  82%|########2 | 2715/3300 [25:38<17:38,  1.81s/it, lr: 1.0e-04 loss: 6.243e-02]
anha:  82%|########2 | 2715/3300 [25:39<17:38,  1.81s/it, lr: 1.0e-04 loss: 4.940e-02]
anha:  82%|########2 | 2715/3300 [25:39<17:38,  1.81s/it, lr: 1.0e-04 loss: 4.940e-02]
anha:  82%|########2 | 2716/3300 [25:39<17:20,  1.78s/it, lr: 1.0e-04 loss: 4.940e-02]
anha:  82%|########2 | 2716/3300 [25:39<17:20,  1.78s/it, lr: 1.0e-04 loss: 4.940e-02]
anha:  82%|########2 | 2716/3300 [25:40<17:20,  1.78s/it, lr: 1.0e-04 loss: 1.091e-01]
anha:  82%|########2 | 2716/3300 [25:40<17:20,  1.78s/it, lr: 1.0e-04 loss: 1.091e-01]
anha:  82%|########2 | 2717/3300 [25:40<15:10,  1.56s/it, lr: 1.0e-04 loss: 1.091e-01]
anha:  82%|########2 | 2717/3300 [25:40<15:10,  1.56s/it, lr: 1.0e-04 loss: 1.091e-01]
anha:  82%|########2 | 2717/3300 [25:41<15:10,  1.56s/it, lr: 1.0e-04 loss: 7.297e-02]
anha:  82%|########2 | 2717/3300 [25:41<15:10,  1.56s/it, lr: 1.0e-04 loss: 7.297e-02]
anha:  82%|########2 | 2718/3300 [25:41<13:32,  1.40s/it, lr: 1.0e-04 loss: 7.297e-02]
anha:  82%|########2 | 2718/3300 [25:41<13:32,  1.40s/it, lr: 1.0e-04 loss: 7.297e-02]
anha:  82%|########2 | 2718/3300 [25:45<13:32,  1.40s/it, lr: 1.0e-04 loss: 3.070e-02]
anha:  82%|########2 | 2718/3300 [25:45<13:32,  1.40s/it, lr: 1.0e-04 loss: 3.070e-02]
anha:  82%|########2 | 2719/3300 [25:45<19:09,  1.98s/it, lr: 1.0e-04 loss: 3.070e-02]
anha:  82%|########2 | 2719/3300 [25:45<19:09,  1.98s/it, lr: 1.0e-04 loss: 3.070e-02]
anha:  82%|########2 | 2719/3300 [25:46<19:09,  1.98s/it, lr: 1.0e-04 loss: 9.816e-02]
anha:  82%|########2 | 2719/3300 [25:46<19:09,  1.98s/it, lr: 1.0e-04 loss: 9.816e-02]
anha:  82%|########2 | 2720/3300 [25:49<19:07,  1.98s/it, lr: 1.0e-04 loss: 1.510e-01]
anha:  82%|########2 | 2720/3300 [25:49<19:07,  1.98s/it, lr: 1.0e-04 loss: 1.510e-01]
anha:  82%|########2 | 2721/3300 [25:49<18:02,  1.87s/it, lr: 1.0e-04 loss: 1.510e-01]
anha:  82%|########2 | 2721/3300 [25:49<18:02,  1.87s/it, lr: 1.0e-04 loss: 1.510e-01]
anha:  82%|########2 | 2721/3300 [25:51<18:02,  1.87s/it, lr: 1.0e-04 loss: 4.225e-02]
anha:  82%|########2 | 2721/3300 [25:51<18:02,  1.87s/it, lr: 1.0e-04 loss: 4.225e-02]
anha:  82%|########2 | 2722/3300 [25:51<18:16,  1.90s/it, lr: 1.0e-04 loss: 4.225e-02]
anha:  82%|########2 | 2722/3300 [25:51<18:16,  1.90s/it, lr: 1.0e-04 loss: 4.225e-02]
anha:  82%|########2 | 2722/3300 [25:55<18:16,  1.90s/it, lr: 1.0e-04 loss: 3.618e-02]
anha:  82%|########2 | 2722/3300 [25:55<18:16,  1.90s/it, lr: 1.0e-04 loss: 3.618e-02]
anha:  83%|########2 | 2723/3300 [25:55<21:59,  2.29s/it, lr: 1.0e-04 loss: 3.618e-02]
anha:  83%|########2 | 2723/3300 [25:55<21:59,  2.29s/it, lr: 1.0e-04 loss: 3.618e-02]
anha:  83%|########2 | 2723/3300 [25:56<21:59,  2.29s/it, lr: 1.0e-04 loss: 1.272e-01]
anha:  83%|########2 | 2723/3300 [25:56<21:59,  2.29s/it, lr: 1.0e-04 loss: 1.272e-01]
anha:  83%|########2 | 2724/3300 [25:56<18:36,  1.94s/it, lr: 1.0e-04 loss: 1.272e-01]
anha:  83%|########2 | 2724/3300 [25:56<18:36,  1.94s/it, lr: 1.0e-04 loss: 1.272e-01]
anha:  83%|########2 | 2724/3300 [25:57<18:36,  1.94s/it, lr: 1.0e-04 loss: 3.466e-02]
anha:  83%|########2 | 2724/3300 [25:57<18:36,  1.94s/it, lr: 1.0e-04 loss: 3.466e-02]
anha:  83%|########2 | 2725/3300 [25:57<18:12,  1.90s/it, lr: 1.0e-04 loss: 3.466e-02]
anha:  83%|########2 | 2725/3300 [25:57<18:12,  1.90s/it, lr: 1.0e-04 loss: 3.466e-02]
anha:  83%|########2 | 2725/3300 [25:58<18:12,  1.90s/it, lr: 1.0e-04 loss: 3.242e-02]
anha:  83%|########2 | 2725/3300 [25:58<18:12,  1.90s/it, lr: 1.0e-04 loss: 3.242e-02]
anha:  83%|########2 | 2726/3300 [25:58<15:43,  1.64s/it, lr: 1.0e-04 loss: 3.242e-02]
anha:  83%|########2 | 2726/3300 [25:58<15:43,  1.64s/it, lr: 1.0e-04 loss: 3.242e-02]
anha:  83%|########2 | 2726/3300 [26:02<15:43,  1.64s/it, lr: 1.0e-04 loss: 3.324e-01]
anha:  83%|########2 | 2726/3300 [26:02<15:43,  1.64s/it, lr: 1.0e-04 loss: 3.324e-01]
anha:  83%|########2 | 2727/3300 [26:02<20:06,  2.11s/it, lr: 1.0e-04 loss: 3.324e-01]
anha:  83%|########2 | 2727/3300 [26:02<20:06,  2.11s/it, lr: 1.0e-04 loss: 3.324e-01]
anha:  83%|########2 | 2727/3300 [26:03<20:06,  2.11s/it, lr: 1.0e-04 loss: 1.248e-01]
anha:  83%|########2 | 2727/3300 [26:03<20:06,  2.11s/it, lr: 1.0e-04 loss: 1.248e-01]
anha:  83%|########2 | 2728/3300 [26:03<17:03,  1.79s/it, lr: 1.0e-04 loss: 1.248e-01]
anha:  83%|########2 | 2728/3300 [26:03<17:03,  1.79s/it, lr: 1.0e-04 loss: 1.248e-01]
anha:  83%|########2 | 2728/3300 [26:05<17:03,  1.79s/it, lr: 1.0e-04 loss: 1.410e-01]
anha:  83%|########2 | 2728/3300 [26:05<17:03,  1.79s/it, lr: 1.0e-04 loss: 1.410e-01]
anha:  83%|########2 | 2729/3300 [26:05<19:02,  2.00s/it, lr: 1.0e-04 loss: 1.410e-01]
anha:  83%|########2 | 2729/3300 [26:05<19:02,  2.00s/it, lr: 1.0e-04 loss: 1.410e-01]
anha:  83%|########2 | 2729/3300 [26:07<19:02,  2.00s/it, lr: 1.0e-04 loss: 5.065e-02]
anha:  83%|########2 | 2729/3300 [26:07<19:02,  2.00s/it, lr: 1.0e-04 loss: 5.065e-02]
anha:  83%|########2 | 2730/3300 [26:11<19:00,  2.00s/it, lr: 1.0e-04 loss: 9.828e-02]
anha:  83%|########2 | 2730/3300 [26:11<19:00,  2.00s/it, lr: 1.0e-04 loss: 9.828e-02]
anha:  83%|########2 | 2731/3300 [26:11<17:51,  1.88s/it, lr: 1.0e-04 loss: 9.828e-02]
anha:  83%|########2 | 2731/3300 [26:11<17:51,  1.88s/it, lr: 1.0e-04 loss: 9.828e-02]
anha:  83%|########2 | 2731/3300 [26:12<17:51,  1.88s/it, lr: 1.0e-04 loss: 7.327e-02]
anha:  83%|########2 | 2731/3300 [26:12<17:51,  1.88s/it, lr: 1.0e-04 loss: 7.327e-02]
anha:  83%|########2 | 2732/3300 [26:12<17:29,  1.85s/it, lr: 1.0e-04 loss: 7.327e-02]
anha:  83%|########2 | 2732/3300 [26:12<17:29,  1.85s/it, lr: 1.0e-04 loss: 7.327e-02]
anha:  83%|########2 | 2732/3300 [26:16<17:29,  1.85s/it, lr: 1.0e-04 loss: 2.909e-02]
anha:  83%|########2 | 2732/3300 [26:16<17:29,  1.85s/it, lr: 1.0e-04 loss: 2.909e-02]
anha:  83%|########2 | 2733/3300 [26:16<21:12,  2.24s/it, lr: 1.0e-04 loss: 2.909e-02]
anha:  83%|########2 | 2733/3300 [26:16<21:12,  2.24s/it, lr: 1.0e-04 loss: 2.909e-02]
anha:  83%|########2 | 2733/3300 [26:17<21:12,  2.24s/it, lr: 1.0e-04 loss: 6.022e-02]
anha:  83%|########2 | 2733/3300 [26:17<21:12,  2.24s/it, lr: 1.0e-04 loss: 6.022e-02]
anha:  83%|########2 | 2734/3300 [26:17<18:15,  1.93s/it, lr: 1.0e-04 loss: 6.022e-02]
anha:  83%|########2 | 2734/3300 [26:17<18:15,  1.93s/it, lr: 1.0e-04 loss: 6.022e-02]
anha:  83%|########2 | 2734/3300 [26:19<18:15,  1.93s/it, lr: 1.0e-04 loss: 5.150e-02]
anha:  83%|########2 | 2734/3300 [26:19<18:15,  1.93s/it, lr: 1.0e-04 loss: 5.150e-02]
anha:  83%|########2 | 2735/3300 [26:19<18:15,  1.94s/it, lr: 1.0e-04 loss: 5.150e-02]
anha:  83%|########2 | 2735/3300 [26:19<18:15,  1.94s/it, lr: 1.0e-04 loss: 5.150e-02]
anha:  83%|########2 | 2735/3300 [26:21<18:15,  1.94s/it, lr: 1.0e-04 loss: 3.709e-02]
anha:  83%|########2 | 2735/3300 [26:21<18:15,  1.94s/it, lr: 1.0e-04 loss: 3.709e-02]
anha:  83%|########2 | 2736/3300 [26:21<18:15,  1.94s/it, lr: 1.0e-04 loss: 3.709e-02]
anha:  83%|########2 | 2736/3300 [26:21<18:15,  1.94s/it, lr: 1.0e-04 loss: 3.709e-02]
anha:  83%|########2 | 2736/3300 [26:22<18:15,  1.94s/it, lr: 1.0e-04 loss: 5.721e-02]
anha:  83%|########2 | 2736/3300 [26:22<18:15,  1.94s/it, lr: 1.0e-04 loss: 5.721e-02]
anha:  83%|########2 | 2737/3300 [26:22<15:41,  1.67s/it, lr: 1.0e-04 loss: 5.721e-02]
anha:  83%|########2 | 2737/3300 [26:22<15:41,  1.67s/it, lr: 1.0e-04 loss: 5.721e-02]
anha:  83%|########2 | 2737/3300 [26:25<15:41,  1.67s/it, lr: 1.0e-04 loss: 6.183e-02]
anha:  83%|########2 | 2737/3300 [26:25<15:41,  1.67s/it, lr: 1.0e-04 loss: 6.183e-02]
anha:  83%|########2 | 2738/3300 [26:25<20:19,  2.17s/it, lr: 1.0e-04 loss: 6.183e-02]
anha:  83%|########2 | 2738/3300 [26:25<20:19,  2.17s/it, lr: 1.0e-04 loss: 6.183e-02]
anha:  83%|########2 | 2738/3300 [26:26<20:19,  2.17s/it, lr: 1.0e-04 loss: 6.417e-02]
anha:  83%|########2 | 2738/3300 [26:26<20:19,  2.17s/it, lr: 1.0e-04 loss: 6.417e-02]
anha:  83%|########2 | 2739/3300 [26:26<17:05,  1.83s/it, lr: 1.0e-04 loss: 6.417e-02]
anha:  83%|########2 | 2739/3300 [26:26<17:05,  1.83s/it, lr: 1.0e-04 loss: 6.417e-02]
anha:  83%|########2 | 2739/3300 [26:30<17:05,  1.83s/it, lr: 1.0e-04 loss: 3.908e-02]
anha:  83%|########2 | 2739/3300 [26:30<17:05,  1.83s/it, lr: 1.0e-04 loss: 3.908e-02]
anha:  83%|########3 | 2740/3300 [26:33<17:03,  1.83s/it, lr: 1.0e-04 loss: 8.418e-02]
anha:  83%|########3 | 2740/3300 [26:33<17:03,  1.83s/it, lr: 1.0e-04 loss: 8.418e-02]
anha:  83%|########3 | 2741/3300 [26:33<16:14,  1.74s/it, lr: 1.0e-04 loss: 8.418e-02]
anha:  83%|########3 | 2741/3300 [26:33<16:14,  1.74s/it, lr: 1.0e-04 loss: 8.418e-02]
anha:  83%|########3 | 2741/3300 [26:34<16:14,  1.74s/it, lr: 1.0e-04 loss: 5.243e-02]
anha:  83%|########3 | 2741/3300 [26:34<16:14,  1.74s/it, lr: 1.0e-04 loss: 5.243e-02]
anha:  83%|########3 | 2742/3300 [26:34<14:30,  1.56s/it, lr: 1.0e-04 loss: 5.243e-02]
anha:  83%|########3 | 2742/3300 [26:34<14:30,  1.56s/it, lr: 1.0e-04 loss: 5.243e-02]
anha:  83%|########3 | 2742/3300 [26:35<14:30,  1.56s/it, lr: 1.0e-04 loss: 5.603e-02]
anha:  83%|########3 | 2742/3300 [26:35<14:30,  1.56s/it, lr: 1.0e-04 loss: 5.603e-02]
anha:  83%|########3 | 2743/3300 [26:35<13:24,  1.44s/it, lr: 1.0e-04 loss: 5.603e-02]
anha:  83%|########3 | 2743/3300 [26:35<13:24,  1.44s/it, lr: 1.0e-04 loss: 5.603e-02]
anha:  83%|########3 | 2743/3300 [26:38<13:24,  1.44s/it, lr: 1.0e-04 loss: 2.323e-01]
anha:  83%|########3 | 2743/3300 [26:38<13:24,  1.44s/it, lr: 1.0e-04 loss: 2.323e-01]
anha:  83%|########3 | 2744/3300 [26:38<18:13,  1.97s/it, lr: 1.0e-04 loss: 2.323e-01]
anha:  83%|########3 | 2744/3300 [26:38<18:13,  1.97s/it, lr: 1.0e-04 loss: 2.323e-01]
anha:  83%|########3 | 2744/3300 [26:40<18:13,  1.97s/it, lr: 1.0e-04 loss: 5.481e-02]
anha:  83%|########3 | 2744/3300 [26:40<18:13,  1.97s/it, lr: 1.0e-04 loss: 5.481e-02]
anha:  83%|########3 | 2745/3300 [26:40<18:12,  1.97s/it, lr: 1.0e-04 loss: 5.481e-02]
anha:  83%|########3 | 2745/3300 [26:40<18:12,  1.97s/it, lr: 1.0e-04 loss: 5.481e-02]
anha:  83%|########3 | 2745/3300 [26:42<18:12,  1.97s/it, lr: 1.0e-04 loss: 5.515e-02]
anha:  83%|########3 | 2745/3300 [26:42<18:12,  1.97s/it, lr: 1.0e-04 loss: 5.515e-02]
anha:  83%|########3 | 2746/3300 [26:42<17:32,  1.90s/it, lr: 1.0e-04 loss: 5.515e-02]
anha:  83%|########3 | 2746/3300 [26:42<17:32,  1.90s/it, lr: 1.0e-04 loss: 5.515e-02]
anha:  83%|########3 | 2746/3300 [26:44<17:32,  1.90s/it, lr: 1.0e-04 loss: 3.171e-02]
anha:  83%|########3 | 2746/3300 [26:44<17:32,  1.90s/it, lr: 1.0e-04 loss: 3.171e-02]
anha:  83%|########3 | 2747/3300 [26:44<17:39,  1.92s/it, lr: 1.0e-04 loss: 3.171e-02]
anha:  83%|########3 | 2747/3300 [26:44<17:39,  1.92s/it, lr: 1.0e-04 loss: 3.171e-02]
anha:  83%|########3 | 2747/3300 [26:46<17:39,  1.92s/it, lr: 1.0e-04 loss: 6.467e-02]
anha:  83%|########3 | 2747/3300 [26:46<17:39,  1.92s/it, lr: 1.0e-04 loss: 6.467e-02]
anha:  83%|########3 | 2748/3300 [26:46<17:10,  1.87s/it, lr: 1.0e-04 loss: 6.467e-02]
anha:  83%|########3 | 2748/3300 [26:46<17:10,  1.87s/it, lr: 1.0e-04 loss: 6.467e-02]
anha:  83%|########3 | 2748/3300 [26:49<17:10,  1.87s/it, lr: 1.0e-04 loss: 6.075e-02]
anha:  83%|########3 | 2748/3300 [26:49<17:10,  1.87s/it, lr: 1.0e-04 loss: 6.075e-02]
anha:  83%|########3 | 2749/3300 [26:49<20:59,  2.29s/it, lr: 1.0e-04 loss: 6.075e-02]
anha:  83%|########3 | 2749/3300 [26:49<20:59,  2.29s/it, lr: 1.0e-04 loss: 6.075e-02]
anha:  83%|########3 | 2749/3300 [26:50<20:59,  2.29s/it, lr: 1.0e-04 loss: 4.613e-02]
anha:  83%|########3 | 2749/3300 [26:50<20:59,  2.29s/it, lr: 1.0e-04 loss: 4.613e-02]


Generating Images: 0it [00:00, ?it/s]
Generating Images: 0it [00:00, ?it/s]


                                     
                                     
anha:  83%|########3 | 2750/3300 [26:52<20:56,  2.29s/it, lr: 1.0e-04 loss: 3.865e-02]
anha:  83%|########3 | 2750/3300 [26:52<20:56,  2.29s/it, lr: 1.0e-04 loss: 3.865e-02]
anha:  83%|########3 | 2751/3300 [26:52<15:28,  1.69s/it, lr: 1.0e-04 loss: 3.865e-02]
anha:  83%|########3 | 2751/3300 [26:52<15:28,  1.69s/it, lr: 1.0e-04 loss: 3.865e-02]
anha:  83%|########3 | 2751/3300 [26:54<15:28,  1.69s/it, lr: 1.0e-04 loss: 5.190e-02]
anha:  83%|########3 | 2751/3300 [26:54<15:28,  1.69s/it, lr: 1.0e-04 loss: 5.190e-02]
anha:  83%|########3 | 2752/3300 [26:54<16:08,  1.77s/it, lr: 1.0e-04 loss: 5.190e-02]
anha:  83%|########3 | 2752/3300 [26:54<16:08,  1.77s/it, lr: 1.0e-04 loss: 5.190e-02]
anha:  83%|########3 | 2752/3300 [26:57<16:08,  1.77s/it, lr: 1.0e-04 loss: 4.219e-02]
anha:  83%|########3 | 2752/3300 [26:57<16:08,  1.77s/it, lr: 1.0e-04 loss: 4.219e-02]
anha:  83%|########3 | 2753/3300 [26:57<20:04,  2.20s/it, lr: 1.0e-04 loss: 4.219e-02]
anha:  83%|########3 | 2753/3300 [26:57<20:04,  2.20s/it, lr: 1.0e-04 loss: 4.219e-02]
anha:  83%|########3 | 2753/3300 [26:59<20:04,  2.20s/it, lr: 1.0e-04 loss: 1.674e-01]
anha:  83%|########3 | 2753/3300 [26:59<20:04,  2.20s/it, lr: 1.0e-04 loss: 1.674e-01]
anha:  83%|########3 | 2754/3300 [26:59<18:53,  2.08s/it, lr: 1.0e-04 loss: 1.674e-01]
anha:  83%|########3 | 2754/3300 [26:59<18:53,  2.08s/it, lr: 1.0e-04 loss: 1.674e-01]
anha:  83%|########3 | 2754/3300 [27:01<18:53,  2.08s/it, lr: 1.0e-04 loss: 5.064e-02]
anha:  83%|########3 | 2754/3300 [27:01<18:53,  2.08s/it, lr: 1.0e-04 loss: 5.064e-02]
anha:  83%|########3 | 2755/3300 [27:01<18:36,  2.05s/it, lr: 1.0e-04 loss: 5.064e-02]
anha:  83%|########3 | 2755/3300 [27:01<18:36,  2.05s/it, lr: 1.0e-04 loss: 5.064e-02]
anha:  83%|########3 | 2755/3300 [27:05<18:36,  2.05s/it, lr: 1.0e-04 loss: 4.489e-02]
anha:  83%|########3 | 2755/3300 [27:05<18:36,  2.05s/it, lr: 1.0e-04 loss: 4.489e-02]
anha:  84%|########3 | 2756/3300 [27:05<21:59,  2.43s/it, lr: 1.0e-04 loss: 4.489e-02]
anha:  84%|########3 | 2756/3300 [27:05<21:59,  2.43s/it, lr: 1.0e-04 loss: 4.489e-02]
anha:  84%|########3 | 2756/3300 [27:08<21:59,  2.43s/it, lr: 1.0e-04 loss: 5.800e-02]
anha:  84%|########3 | 2756/3300 [27:08<21:59,  2.43s/it, lr: 1.0e-04 loss: 5.800e-02]
anha:  84%|########3 | 2757/3300 [27:08<24:46,  2.74s/it, lr: 1.0e-04 loss: 5.800e-02]
anha:  84%|########3 | 2757/3300 [27:08<24:46,  2.74s/it, lr: 1.0e-04 loss: 5.800e-02]
anha:  84%|########3 | 2757/3300 [27:11<24:46,  2.74s/it, lr: 1.0e-04 loss: 3.682e-01]
anha:  84%|########3 | 2757/3300 [27:11<24:46,  2.74s/it, lr: 1.0e-04 loss: 3.682e-01]
anha:  84%|########3 | 2758/3300 [27:11<26:35,  2.94s/it, lr: 1.0e-04 loss: 3.682e-01]
anha:  84%|########3 | 2758/3300 [27:11<26:35,  2.94s/it, lr: 1.0e-04 loss: 3.682e-01]
anha:  84%|########3 | 2758/3300 [27:12<26:35,  2.94s/it, lr: 1.0e-04 loss: 4.135e-02]
anha:  84%|########3 | 2758/3300 [27:12<26:35,  2.94s/it, lr: 1.0e-04 loss: 4.135e-02]
anha:  84%|########3 | 2759/3300 [27:12<21:27,  2.38s/it, lr: 1.0e-04 loss: 4.135e-02]
anha:  84%|########3 | 2759/3300 [27:12<21:27,  2.38s/it, lr: 1.0e-04 loss: 4.135e-02]
anha:  84%|########3 | 2759/3300 [27:16<21:27,  2.38s/it, lr: 1.0e-04 loss: 3.940e-02]
anha:  84%|########3 | 2759/3300 [27:16<21:27,  2.38s/it, lr: 1.0e-04 loss: 3.940e-02]
anha:  84%|########3 | 2760/3300 [27:18<21:24,  2.38s/it, lr: 1.0e-04 loss: 1.015e-01]
anha:  84%|########3 | 2760/3300 [27:18<21:24,  2.38s/it, lr: 1.0e-04 loss: 1.015e-01]
anha:  84%|########3 | 2761/3300 [27:18<15:36,  1.74s/it, lr: 1.0e-04 loss: 1.015e-01]
anha:  84%|########3 | 2761/3300 [27:18<15:36,  1.74s/it, lr: 1.0e-04 loss: 1.015e-01]
anha:  84%|########3 | 2761/3300 [27:19<15:36,  1.74s/it, lr: 1.0e-04 loss: 4.553e-02]
anha:  84%|########3 | 2761/3300 [27:19<15:36,  1.74s/it, lr: 1.0e-04 loss: 4.553e-02]
anha:  84%|########3 | 2762/3300 [27:19<13:55,  1.55s/it, lr: 1.0e-04 loss: 4.553e-02]
anha:  84%|########3 | 2762/3300 [27:19<13:55,  1.55s/it, lr: 1.0e-04 loss: 4.553e-02]
anha:  84%|########3 | 2762/3300 [27:20<13:55,  1.55s/it, lr: 1.0e-04 loss: 6.288e-02]
anha:  84%|########3 | 2762/3300 [27:20<13:55,  1.55s/it, lr: 1.0e-04 loss: 6.288e-02]
anha:  84%|########3 | 2763/3300 [27:20<12:38,  1.41s/it, lr: 1.0e-04 loss: 6.288e-02]
anha:  84%|########3 | 2763/3300 [27:20<12:38,  1.41s/it, lr: 1.0e-04 loss: 6.288e-02]
anha:  84%|########3 | 2763/3300 [27:22<12:38,  1.41s/it, lr: 1.0e-04 loss: 4.215e-02]
anha:  84%|########3 | 2763/3300 [27:22<12:38,  1.41s/it, lr: 1.0e-04 loss: 4.215e-02]
anha:  84%|########3 | 2764/3300 [27:22<13:57,  1.56s/it, lr: 1.0e-04 loss: 4.215e-02]
anha:  84%|########3 | 2764/3300 [27:22<13:57,  1.56s/it, lr: 1.0e-04 loss: 4.215e-02]
anha:  84%|########3 | 2764/3300 [27:24<13:57,  1.56s/it, lr: 1.0e-04 loss: 1.235e-01]
anha:  84%|########3 | 2764/3300 [27:24<13:57,  1.56s/it, lr: 1.0e-04 loss: 1.235e-01]
anha:  84%|########3 | 2765/3300 [27:24<15:00,  1.68s/it, lr: 1.0e-04 loss: 1.235e-01]
anha:  84%|########3 | 2765/3300 [27:24<15:00,  1.68s/it, lr: 1.0e-04 loss: 1.235e-01]
anha:  84%|########3 | 2765/3300 [27:26<15:00,  1.68s/it, lr: 1.0e-04 loss: 6.235e-02]
anha:  84%|########3 | 2765/3300 [27:26<15:00,  1.68s/it, lr: 1.0e-04 loss: 6.235e-02]
anha:  84%|########3 | 2766/3300 [27:26<15:42,  1.76s/it, lr: 1.0e-04 loss: 6.235e-02]
anha:  84%|########3 | 2766/3300 [27:26<15:42,  1.76s/it, lr: 1.0e-04 loss: 6.235e-02]
anha:  84%|########3 | 2766/3300 [27:28<15:42,  1.76s/it, lr: 1.0e-04 loss: 5.525e-02]
anha:  84%|########3 | 2766/3300 [27:28<15:42,  1.76s/it, lr: 1.0e-04 loss: 5.525e-02]
anha:  84%|########3 | 2767/3300 [27:28<16:13,  1.83s/it, lr: 1.0e-04 loss: 5.525e-02]
anha:  84%|########3 | 2767/3300 [27:28<16:13,  1.83s/it, lr: 1.0e-04 loss: 5.525e-02]
anha:  84%|########3 | 2767/3300 [27:29<16:13,  1.83s/it, lr: 1.0e-04 loss: 6.112e-02]
anha:  84%|########3 | 2767/3300 [27:29<16:13,  1.83s/it, lr: 1.0e-04 loss: 6.112e-02]
anha:  84%|########3 | 2768/3300 [27:29<14:05,  1.59s/it, lr: 1.0e-04 loss: 6.112e-02]
anha:  84%|########3 | 2768/3300 [27:29<14:05,  1.59s/it, lr: 1.0e-04 loss: 6.112e-02]
anha:  84%|########3 | 2768/3300 [27:30<14:05,  1.59s/it, lr: 1.0e-04 loss: 3.899e-02]
anha:  84%|########3 | 2768/3300 [27:30<14:05,  1.59s/it, lr: 1.0e-04 loss: 3.899e-02]
anha:  84%|########3 | 2769/3300 [27:30<12:46,  1.44s/it, lr: 1.0e-04 loss: 3.899e-02]
anha:  84%|########3 | 2769/3300 [27:30<12:46,  1.44s/it, lr: 1.0e-04 loss: 3.899e-02]
anha:  84%|########3 | 2769/3300 [27:33<12:46,  1.44s/it, lr: 1.0e-04 loss: 4.833e-02]
anha:  84%|########3 | 2769/3300 [27:33<12:46,  1.44s/it, lr: 1.0e-04 loss: 4.833e-02]
anha:  84%|########3 | 2770/3300 [27:34<12:45,  1.44s/it, lr: 1.0e-04 loss: 1.002e-01]
anha:  84%|########3 | 2770/3300 [27:34<12:45,  1.44s/it, lr: 1.0e-04 loss: 1.002e-01]
anha:  84%|########3 | 2771/3300 [27:34<08:54,  1.01s/it, lr: 1.0e-04 loss: 1.002e-01]
anha:  84%|########3 | 2771/3300 [27:34<08:54,  1.01s/it, lr: 1.0e-04 loss: 1.002e-01]
anha:  84%|########3 | 2771/3300 [27:37<08:54,  1.01s/it, lr: 1.0e-04 loss: 1.272e-01]
anha:  84%|########3 | 2771/3300 [27:37<08:54,  1.01s/it, lr: 1.0e-04 loss: 1.272e-01]
anha:  84%|########4 | 2772/3300 [27:37<13:59,  1.59s/it, lr: 1.0e-04 loss: 1.272e-01]
anha:  84%|########4 | 2772/3300 [27:37<13:59,  1.59s/it, lr: 1.0e-04 loss: 1.272e-01]
anha:  84%|########4 | 2772/3300 [27:39<13:59,  1.59s/it, lr: 1.0e-04 loss: 8.025e-02]
anha:  84%|########4 | 2772/3300 [27:39<13:59,  1.59s/it, lr: 1.0e-04 loss: 8.025e-02]
anha:  84%|########4 | 2773/3300 [27:39<12:40,  1.44s/it, lr: 1.0e-04 loss: 8.025e-02]
anha:  84%|########4 | 2773/3300 [27:39<12:40,  1.44s/it, lr: 1.0e-04 loss: 8.025e-02]
anha:  84%|########4 | 2773/3300 [27:42<12:40,  1.44s/it, lr: 1.0e-04 loss: 1.224e-01]
anha:  84%|########4 | 2773/3300 [27:42<12:40,  1.44s/it, lr: 1.0e-04 loss: 1.224e-01]
anha:  84%|########4 | 2774/3300 [27:42<17:13,  1.97s/it, lr: 1.0e-04 loss: 1.224e-01]
anha:  84%|########4 | 2774/3300 [27:42<17:13,  1.97s/it, lr: 1.0e-04 loss: 1.224e-01]
anha:  84%|########4 | 2774/3300 [27:44<17:13,  1.97s/it, lr: 1.0e-04 loss: 8.406e-02]
anha:  84%|########4 | 2774/3300 [27:44<17:13,  1.97s/it, lr: 1.0e-04 loss: 8.406e-02]
anha:  84%|########4 | 2775/3300 [27:44<16:35,  1.90s/it, lr: 1.0e-04 loss: 8.406e-02]
anha:  84%|########4 | 2775/3300 [27:44<16:35,  1.90s/it, lr: 1.0e-04 loss: 8.406e-02]
anha:  84%|########4 | 2775/3300 [27:45<16:35,  1.90s/it, lr: 1.0e-04 loss: 7.475e-02]
anha:  84%|########4 | 2775/3300 [27:45<16:35,  1.90s/it, lr: 1.0e-04 loss: 7.475e-02]
anha:  84%|########4 | 2776/3300 [27:45<14:21,  1.64s/it, lr: 1.0e-04 loss: 7.475e-02]
anha:  84%|########4 | 2776/3300 [27:45<14:21,  1.64s/it, lr: 1.0e-04 loss: 7.475e-02]
anha:  84%|########4 | 2776/3300 [27:47<14:21,  1.64s/it, lr: 1.0e-04 loss: 7.055e-02]
anha:  84%|########4 | 2776/3300 [27:47<14:21,  1.64s/it, lr: 1.0e-04 loss: 7.055e-02]
anha:  84%|########4 | 2777/3300 [27:47<15:10,  1.74s/it, lr: 1.0e-04 loss: 7.055e-02]
anha:  84%|########4 | 2777/3300 [27:47<15:10,  1.74s/it, lr: 1.0e-04 loss: 7.055e-02]
anha:  84%|########4 | 2777/3300 [27:49<15:10,  1.74s/it, lr: 1.0e-04 loss: 5.825e-02]
anha:  84%|########4 | 2777/3300 [27:49<15:10,  1.74s/it, lr: 1.0e-04 loss: 5.825e-02]
anha:  84%|########4 | 2778/3300 [27:49<15:43,  1.81s/it, lr: 1.0e-04 loss: 5.825e-02]
anha:  84%|########4 | 2778/3300 [27:49<15:43,  1.81s/it, lr: 1.0e-04 loss: 5.825e-02]
anha:  84%|########4 | 2778/3300 [27:51<15:43,  1.81s/it, lr: 1.0e-04 loss: 3.229e-02]
anha:  84%|########4 | 2778/3300 [27:51<15:43,  1.81s/it, lr: 1.0e-04 loss: 3.229e-02]
anha:  84%|########4 | 2779/3300 [27:51<16:21,  1.88s/it, lr: 1.0e-04 loss: 3.229e-02]
anha:  84%|########4 | 2779/3300 [27:51<16:21,  1.88s/it, lr: 1.0e-04 loss: 3.229e-02]
anha:  84%|########4 | 2779/3300 [27:54<16:21,  1.88s/it, lr: 1.0e-04 loss: 7.391e-02]
anha:  84%|########4 | 2779/3300 [27:54<16:21,  1.88s/it, lr: 1.0e-04 loss: 7.391e-02]
anha:  84%|########4 | 2780/3300 [27:56<16:20,  1.88s/it, lr: 1.0e-04 loss: 4.647e-02]
anha:  84%|########4 | 2780/3300 [27:56<16:20,  1.88s/it, lr: 1.0e-04 loss: 4.647e-02]
anha:  84%|########4 | 2781/3300 [27:56<12:44,  1.47s/it, lr: 1.0e-04 loss: 4.647e-02]
anha:  84%|########4 | 2781/3300 [27:56<12:44,  1.47s/it, lr: 1.0e-04 loss: 4.647e-02]
anha:  84%|########4 | 2781/3300 [27:59<12:44,  1.47s/it, lr: 1.0e-04 loss: 4.395e-02]
anha:  84%|########4 | 2781/3300 [27:59<12:44,  1.47s/it, lr: 1.0e-04 loss: 4.395e-02]
anha:  84%|########4 | 2782/3300 [27:59<16:28,  1.91s/it, lr: 1.0e-04 loss: 4.395e-02]
anha:  84%|########4 | 2782/3300 [27:59<16:28,  1.91s/it, lr: 1.0e-04 loss: 4.395e-02]
anha:  84%|########4 | 2782/3300 [28:00<16:28,  1.91s/it, lr: 1.0e-04 loss: 7.461e-02]
anha:  84%|########4 | 2782/3300 [28:00<16:28,  1.91s/it, lr: 1.0e-04 loss: 7.461e-02]
anha:  84%|########4 | 2783/3300 [28:00<14:23,  1.67s/it, lr: 1.0e-04 loss: 7.461e-02]
anha:  84%|########4 | 2783/3300 [28:00<14:23,  1.67s/it, lr: 1.0e-04 loss: 7.461e-02]
anha:  84%|########4 | 2783/3300 [28:02<14:23,  1.67s/it, lr: 1.0e-04 loss: 6.361e-02]
anha:  84%|########4 | 2783/3300 [28:02<14:23,  1.67s/it, lr: 1.0e-04 loss: 6.361e-02]
anha:  84%|########4 | 2784/3300 [28:02<15:01,  1.75s/it, lr: 1.0e-04 loss: 6.361e-02]
anha:  84%|########4 | 2784/3300 [28:02<15:01,  1.75s/it, lr: 1.0e-04 loss: 6.361e-02]
anha:  84%|########4 | 2784/3300 [28:04<15:01,  1.75s/it, lr: 1.0e-04 loss: 7.058e-02]
anha:  84%|########4 | 2784/3300 [28:04<15:01,  1.75s/it, lr: 1.0e-04 loss: 7.058e-02]
anha:  84%|########4 | 2785/3300 [28:04<15:32,  1.81s/it, lr: 1.0e-04 loss: 7.058e-02]
anha:  84%|########4 | 2785/3300 [28:04<15:32,  1.81s/it, lr: 1.0e-04 loss: 7.058e-02]
anha:  84%|########4 | 2785/3300 [28:06<15:32,  1.81s/it, lr: 1.0e-04 loss: 3.954e-02]
anha:  84%|########4 | 2785/3300 [28:06<15:32,  1.81s/it, lr: 1.0e-04 loss: 3.954e-02]
anha:  84%|########4 | 2786/3300 [28:06<15:51,  1.85s/it, lr: 1.0e-04 loss: 3.954e-02]
anha:  84%|########4 | 2786/3300 [28:06<15:51,  1.85s/it, lr: 1.0e-04 loss: 3.954e-02]
anha:  84%|########4 | 2786/3300 [28:07<15:51,  1.85s/it, lr: 1.0e-04 loss: 6.565e-02]
anha:  84%|########4 | 2786/3300 [28:07<15:51,  1.85s/it, lr: 1.0e-04 loss: 6.565e-02]
anha:  84%|########4 | 2787/3300 [28:07<13:41,  1.60s/it, lr: 1.0e-04 loss: 6.565e-02]
anha:  84%|########4 | 2787/3300 [28:07<13:41,  1.60s/it, lr: 1.0e-04 loss: 6.565e-02]
anha:  84%|########4 | 2787/3300 [28:09<13:41,  1.60s/it, lr: 1.0e-04 loss: 3.892e-02]
anha:  84%|########4 | 2787/3300 [28:09<13:41,  1.60s/it, lr: 1.0e-04 loss: 3.892e-02]
anha:  84%|########4 | 2788/3300 [28:09<14:47,  1.73s/it, lr: 1.0e-04 loss: 3.892e-02]
anha:  84%|########4 | 2788/3300 [28:09<14:47,  1.73s/it, lr: 1.0e-04 loss: 3.892e-02]
anha:  84%|########4 | 2788/3300 [28:13<14:47,  1.73s/it, lr: 1.0e-04 loss: 9.861e-02]
anha:  84%|########4 | 2788/3300 [28:13<14:47,  1.73s/it, lr: 1.0e-04 loss: 9.861e-02]
anha:  85%|########4 | 2789/3300 [28:13<19:12,  2.26s/it, lr: 1.0e-04 loss: 9.861e-02]
anha:  85%|########4 | 2789/3300 [28:13<19:12,  2.26s/it, lr: 1.0e-04 loss: 9.861e-02]
anha:  85%|########4 | 2789/3300 [28:14<19:12,  2.26s/it, lr: 1.0e-04 loss: 4.796e-02]
anha:  85%|########4 | 2789/3300 [28:14<19:12,  2.26s/it, lr: 1.0e-04 loss: 4.796e-02]
anha:  85%|########4 | 2790/3300 [28:17<19:10,  2.26s/it, lr: 1.0e-04 loss: 1.249e-01]
anha:  85%|########4 | 2790/3300 [28:17<19:10,  2.26s/it, lr: 1.0e-04 loss: 1.249e-01]
anha:  85%|########4 | 2791/3300 [28:17<16:55,  1.99s/it, lr: 1.0e-04 loss: 1.249e-01]
anha:  85%|########4 | 2791/3300 [28:17<16:55,  1.99s/it, lr: 1.0e-04 loss: 1.249e-01]
anha:  85%|########4 | 2791/3300 [28:18<16:55,  1.99s/it, lr: 1.0e-04 loss: 3.854e-02]
anha:  85%|########4 | 2791/3300 [28:18<16:55,  1.99s/it, lr: 1.0e-04 loss: 3.854e-02]
anha:  85%|########4 | 2792/3300 [28:18<14:47,  1.75s/it, lr: 1.0e-04 loss: 3.854e-02]
anha:  85%|########4 | 2792/3300 [28:18<14:47,  1.75s/it, lr: 1.0e-04 loss: 3.854e-02]
anha:  85%|########4 | 2792/3300 [28:19<14:47,  1.75s/it, lr: 1.0e-04 loss: 2.178e-01]
anha:  85%|########4 | 2792/3300 [28:19<14:47,  1.75s/it, lr: 1.0e-04 loss: 2.178e-01]
anha:  85%|########4 | 2793/3300 [28:19<13:06,  1.55s/it, lr: 1.0e-04 loss: 2.178e-01]
anha:  85%|########4 | 2793/3300 [28:19<13:06,  1.55s/it, lr: 1.0e-04 loss: 2.178e-01]
anha:  85%|########4 | 2793/3300 [28:22<13:06,  1.55s/it, lr: 1.0e-04 loss: 7.544e-02]
anha:  85%|########4 | 2793/3300 [28:22<13:06,  1.55s/it, lr: 1.0e-04 loss: 7.544e-02]
anha:  85%|########4 | 2794/3300 [28:22<17:28,  2.07s/it, lr: 1.0e-04 loss: 7.544e-02]
anha:  85%|########4 | 2794/3300 [28:22<17:28,  2.07s/it, lr: 1.0e-04 loss: 7.544e-02]
anha:  85%|########4 | 2794/3300 [28:23<17:28,  2.07s/it, lr: 1.0e-04 loss: 3.766e-02]
anha:  85%|########4 | 2794/3300 [28:23<17:28,  2.07s/it, lr: 1.0e-04 loss: 3.766e-02]
anha:  85%|########4 | 2795/3300 [28:23<14:56,  1.78s/it, lr: 1.0e-04 loss: 3.766e-02]
anha:  85%|########4 | 2795/3300 [28:23<14:56,  1.78s/it, lr: 1.0e-04 loss: 3.766e-02]
anha:  85%|########4 | 2795/3300 [28:27<14:56,  1.78s/it, lr: 1.0e-04 loss: 1.246e-01]
anha:  85%|########4 | 2795/3300 [28:27<14:56,  1.78s/it, lr: 1.0e-04 loss: 1.246e-01]
anha:  85%|########4 | 2796/3300 [28:27<19:00,  2.26s/it, lr: 1.0e-04 loss: 1.246e-01]
anha:  85%|########4 | 2796/3300 [28:27<19:00,  2.26s/it, lr: 1.0e-04 loss: 1.246e-01]
anha:  85%|########4 | 2796/3300 [28:30<19:00,  2.26s/it, lr: 1.0e-04 loss: 4.559e-02]
anha:  85%|########4 | 2796/3300 [28:30<19:00,  2.26s/it, lr: 1.0e-04 loss: 4.559e-02]
anha:  85%|########4 | 2797/3300 [28:30<21:54,  2.61s/it, lr: 1.0e-04 loss: 4.559e-02]
anha:  85%|########4 | 2797/3300 [28:30<21:54,  2.61s/it, lr: 1.0e-04 loss: 4.559e-02]
anha:  85%|########4 | 2797/3300 [28:32<21:54,  2.61s/it, lr: 1.0e-04 loss: 5.157e-02]
anha:  85%|########4 | 2797/3300 [28:32<21:54,  2.61s/it, lr: 1.0e-04 loss: 5.157e-02]
anha:  85%|########4 | 2798/3300 [28:32<19:42,  2.36s/it, lr: 1.0e-04 loss: 5.157e-02]
anha:  85%|########4 | 2798/3300 [28:32<19:42,  2.36s/it, lr: 1.0e-04 loss: 5.157e-02]
anha:  85%|########4 | 2798/3300 [28:35<19:42,  2.36s/it, lr: 1.0e-04 loss: 4.523e-02]
anha:  85%|########4 | 2798/3300 [28:35<19:42,  2.36s/it, lr: 1.0e-04 loss: 4.523e-02]
anha:  85%|########4 | 2799/3300 [28:35<21:57,  2.63s/it, lr: 1.0e-04 loss: 4.523e-02]
anha:  85%|########4 | 2799/3300 [28:35<21:57,  2.63s/it, lr: 1.0e-04 loss: 4.523e-02]
anha:  85%|########4 | 2799/3300 [28:37<21:57,  2.63s/it, lr: 1.0e-04 loss: 1.820e-01]
anha:  85%|########4 | 2799/3300 [28:37<21:57,  2.63s/it, lr: 1.0e-04 loss: 1.820e-01]
anha:  85%|########4 | 2800/3300 [28:38<21:55,  2.63s/it, lr: 1.0e-04 loss: 5.866e-02]
anha:  85%|########4 | 2800/3300 [28:38<21:55,  2.63s/it, lr: 1.0e-04 loss: 5.866e-02]
anha:  85%|########4 | 2801/3300 [28:38<13:45,  1.65s/it, lr: 1.0e-04 loss: 5.866e-02]
anha:  85%|########4 | 2801/3300 [28:38<13:45,  1.65s/it, lr: 1.0e-04 loss: 5.866e-02]
anha:  85%|########4 | 2801/3300 [28:40<13:45,  1.65s/it, lr: 1.0e-04 loss: 4.895e-02]
anha:  85%|########4 | 2801/3300 [28:40<13:45,  1.65s/it, lr: 1.0e-04 loss: 4.895e-02]
anha:  85%|########4 | 2802/3300 [28:40<14:20,  1.73s/it, lr: 1.0e-04 loss: 4.895e-02]
anha:  85%|########4 | 2802/3300 [28:40<14:20,  1.73s/it, lr: 1.0e-04 loss: 4.895e-02]
anha:  85%|########4 | 2802/3300 [28:42<14:20,  1.73s/it, lr: 1.0e-04 loss: 5.687e-02]
anha:  85%|########4 | 2802/3300 [28:42<14:20,  1.73s/it, lr: 1.0e-04 loss: 5.687e-02]
anha:  85%|########4 | 2803/3300 [28:42<14:19,  1.73s/it, lr: 1.0e-04 loss: 5.687e-02]
anha:  85%|########4 | 2803/3300 [28:42<14:19,  1.73s/it, lr: 1.0e-04 loss: 5.687e-02]
anha:  85%|########4 | 2803/3300 [28:44<14:19,  1.73s/it, lr: 1.0e-04 loss: 8.031e-02]
anha:  85%|########4 | 2803/3300 [28:44<14:19,  1.73s/it, lr: 1.0e-04 loss: 8.031e-02]
anha:  85%|########4 | 2804/3300 [28:44<14:15,  1.73s/it, lr: 1.0e-04 loss: 8.031e-02]
anha:  85%|########4 | 2804/3300 [28:44<14:15,  1.73s/it, lr: 1.0e-04 loss: 8.031e-02]
anha:  85%|########4 | 2804/3300 [28:45<14:15,  1.73s/it, lr: 1.0e-04 loss: 6.638e-02]
anha:  85%|########4 | 2804/3300 [28:45<14:15,  1.73s/it, lr: 1.0e-04 loss: 6.638e-02]
anha:  85%|########5 | 2805/3300 [28:45<12:34,  1.52s/it, lr: 1.0e-04 loss: 6.638e-02]
anha:  85%|########5 | 2805/3300 [28:45<12:34,  1.52s/it, lr: 1.0e-04 loss: 6.638e-02]
anha:  85%|########5 | 2805/3300 [28:46<12:34,  1.52s/it, lr: 1.0e-04 loss: 5.810e-02]
anha:  85%|########5 | 2805/3300 [28:46<12:34,  1.52s/it, lr: 1.0e-04 loss: 5.810e-02]
anha:  85%|########5 | 2806/3300 [28:46<11:19,  1.38s/it, lr: 1.0e-04 loss: 5.810e-02]
anha:  85%|########5 | 2806/3300 [28:46<11:19,  1.38s/it, lr: 1.0e-04 loss: 5.810e-02]
anha:  85%|########5 | 2806/3300 [28:48<11:19,  1.38s/it, lr: 1.0e-04 loss: 6.005e-02]
anha:  85%|########5 | 2806/3300 [28:48<11:19,  1.38s/it, lr: 1.0e-04 loss: 6.005e-02]
anha:  85%|########5 | 2807/3300 [28:48<12:50,  1.56s/it, lr: 1.0e-04 loss: 6.005e-02]
anha:  85%|########5 | 2807/3300 [28:48<12:50,  1.56s/it, lr: 1.0e-04 loss: 6.005e-02]
anha:  85%|########5 | 2807/3300 [28:49<12:50,  1.56s/it, lr: 1.0e-04 loss: 6.499e-02]
anha:  85%|########5 | 2807/3300 [28:49<12:50,  1.56s/it, lr: 1.0e-04 loss: 6.499e-02]
anha:  85%|########5 | 2808/3300 [28:49<11:29,  1.40s/it, lr: 1.0e-04 loss: 6.499e-02]
anha:  85%|########5 | 2808/3300 [28:49<11:29,  1.40s/it, lr: 1.0e-04 loss: 6.499e-02]
anha:  85%|########5 | 2808/3300 [28:50<11:29,  1.40s/it, lr: 1.0e-04 loss: 1.417e-01]
anha:  85%|########5 | 2808/3300 [28:50<11:29,  1.40s/it, lr: 1.0e-04 loss: 1.417e-01]
anha:  85%|########5 | 2809/3300 [28:50<10:28,  1.28s/it, lr: 1.0e-04 loss: 1.417e-01]
anha:  85%|########5 | 2809/3300 [28:50<10:28,  1.28s/it, lr: 1.0e-04 loss: 1.417e-01]
anha:  85%|########5 | 2809/3300 [28:51<10:28,  1.28s/it, lr: 1.0e-04 loss: 7.340e-02]
anha:  85%|########5 | 2809/3300 [28:51<10:28,  1.28s/it, lr: 1.0e-04 loss: 7.340e-02]
anha:  85%|########5 | 2810/3300 [28:52<10:27,  1.28s/it, lr: 1.0e-04 loss: 6.535e-02]
anha:  85%|########5 | 2810/3300 [28:52<10:27,  1.28s/it, lr: 1.0e-04 loss: 6.535e-02]
anha:  85%|########5 | 2811/3300 [28:52<07:32,  1.08it/s, lr: 1.0e-04 loss: 6.535e-02]
anha:  85%|########5 | 2811/3300 [28:52<07:32,  1.08it/s, lr: 1.0e-04 loss: 6.535e-02]
anha:  85%|########5 | 2811/3300 [28:53<07:32,  1.08it/s, lr: 1.0e-04 loss: 1.128e-01]
anha:  85%|########5 | 2811/3300 [28:53<07:32,  1.08it/s, lr: 1.0e-04 loss: 1.128e-01]
anha:  85%|########5 | 2812/3300 [28:53<09:08,  1.12s/it, lr: 1.0e-04 loss: 1.128e-01]
anha:  85%|########5 | 2812/3300 [28:53<09:08,  1.12s/it, lr: 1.0e-04 loss: 1.128e-01]
anha:  85%|########5 | 2812/3300 [28:57<09:08,  1.12s/it, lr: 1.0e-04 loss: 3.293e-02]
anha:  85%|########5 | 2812/3300 [28:57<09:08,  1.12s/it, lr: 1.0e-04 loss: 3.293e-02]
anha:  85%|########5 | 2813/3300 [28:57<13:51,  1.71s/it, lr: 1.0e-04 loss: 3.293e-02]
anha:  85%|########5 | 2813/3300 [28:57<13:51,  1.71s/it, lr: 1.0e-04 loss: 3.293e-02]
anha:  85%|########5 | 2813/3300 [28:59<13:51,  1.71s/it, lr: 1.0e-04 loss: 1.433e-01]
anha:  85%|########5 | 2813/3300 [28:59<13:51,  1.71s/it, lr: 1.0e-04 loss: 1.433e-01]
anha:  85%|########5 | 2814/3300 [28:59<14:28,  1.79s/it, lr: 1.0e-04 loss: 1.433e-01]
anha:  85%|########5 | 2814/3300 [28:59<14:28,  1.79s/it, lr: 1.0e-04 loss: 1.433e-01]
anha:  85%|########5 | 2814/3300 [29:01<14:28,  1.79s/it, lr: 1.0e-04 loss: 5.212e-02]
anha:  85%|########5 | 2814/3300 [29:01<14:28,  1.79s/it, lr: 1.0e-04 loss: 5.212e-02]
anha:  85%|########5 | 2815/3300 [29:01<14:52,  1.84s/it, lr: 1.0e-04 loss: 5.212e-02]
anha:  85%|########5 | 2815/3300 [29:01<14:52,  1.84s/it, lr: 1.0e-04 loss: 5.212e-02]
anha:  85%|########5 | 2815/3300 [29:03<14:52,  1.84s/it, lr: 1.0e-04 loss: 5.810e-02]
anha:  85%|########5 | 2815/3300 [29:03<14:52,  1.84s/it, lr: 1.0e-04 loss: 5.810e-02]
anha:  85%|########5 | 2816/3300 [29:03<15:18,  1.90s/it, lr: 1.0e-04 loss: 5.810e-02]
anha:  85%|########5 | 2816/3300 [29:03<15:18,  1.90s/it, lr: 1.0e-04 loss: 5.810e-02]
anha:  85%|########5 | 2816/3300 [29:06<15:18,  1.90s/it, lr: 1.0e-04 loss: 2.520e-01]
anha:  85%|########5 | 2816/3300 [29:06<15:18,  1.90s/it, lr: 1.0e-04 loss: 2.520e-01]
anha:  85%|########5 | 2817/3300 [29:06<18:31,  2.30s/it, lr: 1.0e-04 loss: 2.520e-01]
anha:  85%|########5 | 2817/3300 [29:06<18:31,  2.30s/it, lr: 1.0e-04 loss: 2.520e-01]
anha:  85%|########5 | 2817/3300 [29:07<18:31,  2.30s/it, lr: 1.0e-04 loss: 5.932e-02]
anha:  85%|########5 | 2817/3300 [29:07<18:31,  2.30s/it, lr: 1.0e-04 loss: 5.932e-02]
anha:  85%|########5 | 2818/3300 [29:07<15:28,  1.93s/it, lr: 1.0e-04 loss: 5.932e-02]
anha:  85%|########5 | 2818/3300 [29:07<15:28,  1.93s/it, lr: 1.0e-04 loss: 5.932e-02]
anha:  85%|########5 | 2818/3300 [29:11<15:28,  1.93s/it, lr: 1.0e-04 loss: 3.933e-02]
anha:  85%|########5 | 2818/3300 [29:11<15:28,  1.93s/it, lr: 1.0e-04 loss: 3.933e-02]
anha:  85%|########5 | 2819/3300 [29:11<19:07,  2.38s/it, lr: 1.0e-04 loss: 3.933e-02]
anha:  85%|########5 | 2819/3300 [29:11<19:07,  2.38s/it, lr: 1.0e-04 loss: 3.933e-02]
anha:  85%|########5 | 2819/3300 [29:12<19:07,  2.38s/it, lr: 1.0e-04 loss: 3.862e-02]
anha:  85%|########5 | 2819/3300 [29:12<19:07,  2.38s/it, lr: 1.0e-04 loss: 3.862e-02]
anha:  85%|########5 | 2820/3300 [29:14<19:04,  2.38s/it, lr: 1.0e-04 loss: 1.896e-01]
anha:  85%|########5 | 2820/3300 [29:14<19:04,  2.38s/it, lr: 1.0e-04 loss: 1.896e-01]
anha:  85%|########5 | 2821/3300 [29:14<13:55,  1.74s/it, lr: 1.0e-04 loss: 1.896e-01]
anha:  85%|########5 | 2821/3300 [29:14<13:55,  1.74s/it, lr: 1.0e-04 loss: 1.896e-01]
anha:  85%|########5 | 2821/3300 [29:18<13:55,  1.74s/it, lr: 1.0e-04 loss: 2.449e-01]
anha:  85%|########5 | 2821/3300 [29:18<13:55,  1.74s/it, lr: 1.0e-04 loss: 2.449e-01]
anha:  86%|########5 | 2822/3300 [29:18<17:07,  2.15s/it, lr: 1.0e-04 loss: 2.449e-01]
anha:  86%|########5 | 2822/3300 [29:18<17:07,  2.15s/it, lr: 1.0e-04 loss: 2.449e-01]
anha:  86%|########5 | 2822/3300 [29:21<17:07,  2.15s/it, lr: 1.0e-04 loss: 9.893e-02]
anha:  86%|########5 | 2822/3300 [29:21<17:07,  2.15s/it, lr: 1.0e-04 loss: 9.893e-02]
anha:  86%|########5 | 2823/3300 [29:21<19:53,  2.50s/it, lr: 1.0e-04 loss: 9.893e-02]
anha:  86%|########5 | 2823/3300 [29:21<19:53,  2.50s/it, lr: 1.0e-04 loss: 9.893e-02]
anha:  86%|########5 | 2823/3300 [29:24<19:53,  2.50s/it, lr: 1.0e-04 loss: 5.216e-02]
anha:  86%|########5 | 2823/3300 [29:24<19:53,  2.50s/it, lr: 1.0e-04 loss: 5.216e-02]
anha:  86%|########5 | 2824/3300 [29:24<21:43,  2.74s/it, lr: 1.0e-04 loss: 5.216e-02]
anha:  86%|########5 | 2824/3300 [29:24<21:43,  2.74s/it, lr: 1.0e-04 loss: 5.216e-02]
anha:  86%|########5 | 2824/3300 [29:25<21:43,  2.74s/it, lr: 1.0e-04 loss: 6.641e-02]
anha:  86%|########5 | 2824/3300 [29:25<21:43,  2.74s/it, lr: 1.0e-04 loss: 6.641e-02]
anha:  86%|########5 | 2825/3300 [29:25<17:51,  2.26s/it, lr: 1.0e-04 loss: 6.641e-02]
anha:  86%|########5 | 2825/3300 [29:25<17:51,  2.26s/it, lr: 1.0e-04 loss: 6.641e-02]
anha:  86%|########5 | 2825/3300 [29:28<17:51,  2.26s/it, lr: 1.0e-04 loss: 3.446e-02]
anha:  86%|########5 | 2825/3300 [29:28<17:51,  2.26s/it, lr: 1.0e-04 loss: 3.446e-02]
anha:  86%|########5 | 2826/3300 [29:28<17:18,  2.19s/it, lr: 1.0e-04 loss: 3.446e-02]
anha:  86%|########5 | 2826/3300 [29:28<17:18,  2.19s/it, lr: 1.0e-04 loss: 3.446e-02]
anha:  86%|########5 | 2826/3300 [29:29<17:18,  2.19s/it, lr: 1.0e-04 loss: 3.555e-02]
anha:  86%|########5 | 2826/3300 [29:29<17:18,  2.19s/it, lr: 1.0e-04 loss: 3.555e-02]
anha:  86%|########5 | 2827/3300 [29:29<14:34,  1.85s/it, lr: 1.0e-04 loss: 3.555e-02]
anha:  86%|########5 | 2827/3300 [29:29<14:34,  1.85s/it, lr: 1.0e-04 loss: 3.555e-02]
anha:  86%|########5 | 2827/3300 [29:32<14:34,  1.85s/it, lr: 1.0e-04 loss: 3.695e-01]
anha:  86%|########5 | 2827/3300 [29:32<14:34,  1.85s/it, lr: 1.0e-04 loss: 3.695e-01]
anha:  86%|########5 | 2828/3300 [29:32<17:33,  2.23s/it, lr: 1.0e-04 loss: 3.695e-01]
anha:  86%|########5 | 2828/3300 [29:32<17:33,  2.23s/it, lr: 1.0e-04 loss: 3.695e-01]
anha:  86%|########5 | 2828/3300 [29:35<17:33,  2.23s/it, lr: 1.0e-04 loss: 6.693e-02]
anha:  86%|########5 | 2828/3300 [29:35<17:33,  2.23s/it, lr: 1.0e-04 loss: 6.693e-02]
anha:  86%|########5 | 2829/3300 [29:35<19:57,  2.54s/it, lr: 1.0e-04 loss: 6.693e-02]
anha:  86%|########5 | 2829/3300 [29:35<19:57,  2.54s/it, lr: 1.0e-04 loss: 6.693e-02]
anha:  86%|########5 | 2829/3300 [29:36<19:57,  2.54s/it, lr: 1.0e-04 loss: 4.954e-02]
anha:  86%|########5 | 2829/3300 [29:36<19:57,  2.54s/it, lr: 1.0e-04 loss: 4.954e-02]
anha:  86%|########5 | 2830/3300 [29:39<19:55,  2.54s/it, lr: 1.0e-04 loss: 7.362e-02]
anha:  86%|########5 | 2830/3300 [29:39<19:55,  2.54s/it, lr: 1.0e-04 loss: 7.362e-02]
anha:  86%|########5 | 2831/3300 [29:39<17:00,  2.18s/it, lr: 1.0e-04 loss: 7.362e-02]
anha:  86%|########5 | 2831/3300 [29:39<17:00,  2.18s/it, lr: 1.0e-04 loss: 7.362e-02]
anha:  86%|########5 | 2831/3300 [29:40<17:00,  2.18s/it, lr: 1.0e-04 loss: 6.993e-02]
anha:  86%|########5 | 2831/3300 [29:40<17:00,  2.18s/it, lr: 1.0e-04 loss: 6.993e-02]
anha:  86%|########5 | 2832/3300 [29:40<14:43,  1.89s/it, lr: 1.0e-04 loss: 6.993e-02]
anha:  86%|########5 | 2832/3300 [29:40<14:43,  1.89s/it, lr: 1.0e-04 loss: 6.993e-02]
anha:  86%|########5 | 2832/3300 [29:44<14:43,  1.89s/it, lr: 1.0e-04 loss: 7.197e-02]
anha:  86%|########5 | 2832/3300 [29:44<14:43,  1.89s/it, lr: 1.0e-04 loss: 7.197e-02]
anha:  86%|########5 | 2833/3300 [29:44<17:50,  2.29s/it, lr: 1.0e-04 loss: 7.197e-02]
anha:  86%|########5 | 2833/3300 [29:44<17:50,  2.29s/it, lr: 1.0e-04 loss: 7.197e-02]
anha:  86%|########5 | 2833/3300 [29:45<17:50,  2.29s/it, lr: 1.0e-04 loss: 4.502e-02]
anha:  86%|########5 | 2833/3300 [29:45<17:50,  2.29s/it, lr: 1.0e-04 loss: 4.502e-02]
anha:  86%|########5 | 2834/3300 [29:45<15:06,  1.94s/it, lr: 1.0e-04 loss: 4.502e-02]
anha:  86%|########5 | 2834/3300 [29:45<15:06,  1.94s/it, lr: 1.0e-04 loss: 4.502e-02]
anha:  86%|########5 | 2834/3300 [29:46<15:06,  1.94s/it, lr: 1.0e-04 loss: 4.396e-02]
anha:  86%|########5 | 2834/3300 [29:46<15:06,  1.94s/it, lr: 1.0e-04 loss: 4.396e-02]
anha:  86%|########5 | 2835/3300 [29:46<13:18,  1.72s/it, lr: 1.0e-04 loss: 4.396e-02]
anha:  86%|########5 | 2835/3300 [29:46<13:18,  1.72s/it, lr: 1.0e-04 loss: 4.396e-02]
anha:  86%|########5 | 2835/3300 [29:49<13:18,  1.72s/it, lr: 1.0e-04 loss: 3.138e-02]
anha:  86%|########5 | 2835/3300 [29:49<13:18,  1.72s/it, lr: 1.0e-04 loss: 3.138e-02]
anha:  86%|########5 | 2836/3300 [29:49<15:01,  1.94s/it, lr: 1.0e-04 loss: 3.138e-02]
anha:  86%|########5 | 2836/3300 [29:49<15:01,  1.94s/it, lr: 1.0e-04 loss: 3.138e-02]
anha:  86%|########5 | 2836/3300 [29:52<15:01,  1.94s/it, lr: 1.0e-04 loss: 7.601e-02]
anha:  86%|########5 | 2836/3300 [29:52<15:01,  1.94s/it, lr: 1.0e-04 loss: 7.601e-02]
anha:  86%|########5 | 2837/3300 [29:52<17:52,  2.32s/it, lr: 1.0e-04 loss: 7.601e-02]
anha:  86%|########5 | 2837/3300 [29:52<17:52,  2.32s/it, lr: 1.0e-04 loss: 7.601e-02]
anha:  86%|########5 | 2837/3300 [29:55<17:52,  2.32s/it, lr: 1.0e-04 loss: 9.657e-02]
anha:  86%|########5 | 2837/3300 [29:55<17:52,  2.32s/it, lr: 1.0e-04 loss: 9.657e-02]
anha:  86%|########6 | 2838/3300 [29:55<20:21,  2.64s/it, lr: 1.0e-04 loss: 9.657e-02]
anha:  86%|########6 | 2838/3300 [29:55<20:21,  2.64s/it, lr: 1.0e-04 loss: 9.657e-02]
anha:  86%|########6 | 2838/3300 [29:57<20:21,  2.64s/it, lr: 1.0e-04 loss: 9.514e-02]
anha:  86%|########6 | 2838/3300 [29:57<20:21,  2.64s/it, lr: 1.0e-04 loss: 9.514e-02]
anha:  86%|########6 | 2839/3300 [29:57<18:13,  2.37s/it, lr: 1.0e-04 loss: 9.514e-02]
anha:  86%|########6 | 2839/3300 [29:57<18:13,  2.37s/it, lr: 1.0e-04 loss: 9.514e-02]
anha:  86%|########6 | 2839/3300 [29:59<18:13,  2.37s/it, lr: 1.0e-04 loss: 6.758e-02]
anha:  86%|########6 | 2839/3300 [29:59<18:13,  2.37s/it, lr: 1.0e-04 loss: 6.758e-02]
anha:  86%|########6 | 2840/3300 [30:00<18:10,  2.37s/it, lr: 1.0e-04 loss: 5.785e-02]
anha:  86%|########6 | 2840/3300 [30:00<18:10,  2.37s/it, lr: 1.0e-04 loss: 5.785e-02]
anha:  86%|########6 | 2841/3300 [30:00<11:36,  1.52s/it, lr: 1.0e-04 loss: 5.785e-02]
anha:  86%|########6 | 2841/3300 [30:00<11:36,  1.52s/it, lr: 1.0e-04 loss: 5.785e-02]
anha:  86%|########6 | 2841/3300 [30:03<11:36,  1.52s/it, lr: 1.0e-04 loss: 1.778e-01]
anha:  86%|########6 | 2841/3300 [30:03<11:36,  1.52s/it, lr: 1.0e-04 loss: 1.778e-01]
anha:  86%|########6 | 2842/3300 [30:03<15:16,  2.00s/it, lr: 1.0e-04 loss: 1.778e-01]
anha:  86%|########6 | 2842/3300 [30:03<15:16,  2.00s/it, lr: 1.0e-04 loss: 1.778e-01]
anha:  86%|########6 | 2842/3300 [30:04<15:16,  2.00s/it, lr: 1.0e-04 loss: 4.702e-02]
anha:  86%|########6 | 2842/3300 [30:04<15:16,  2.00s/it, lr: 1.0e-04 loss: 4.702e-02]
anha:  86%|########6 | 2843/3300 [30:04<13:16,  1.74s/it, lr: 1.0e-04 loss: 4.702e-02]
anha:  86%|########6 | 2843/3300 [30:04<13:16,  1.74s/it, lr: 1.0e-04 loss: 4.702e-02]
anha:  86%|########6 | 2843/3300 [30:06<13:16,  1.74s/it, lr: 1.0e-04 loss: 9.069e-02]
anha:  86%|########6 | 2843/3300 [30:06<13:16,  1.74s/it, lr: 1.0e-04 loss: 9.069e-02]
anha:  86%|########6 | 2844/3300 [30:06<12:03,  1.59s/it, lr: 1.0e-04 loss: 9.069e-02]
anha:  86%|########6 | 2844/3300 [30:06<12:03,  1.59s/it, lr: 1.0e-04 loss: 9.069e-02]
anha:  86%|########6 | 2844/3300 [30:08<12:03,  1.59s/it, lr: 1.0e-04 loss: 5.189e-02]
anha:  86%|########6 | 2844/3300 [30:08<12:03,  1.59s/it, lr: 1.0e-04 loss: 5.189e-02]
anha:  86%|########6 | 2845/3300 [30:08<12:50,  1.69s/it, lr: 1.0e-04 loss: 5.189e-02]
anha:  86%|########6 | 2845/3300 [30:08<12:50,  1.69s/it, lr: 1.0e-04 loss: 5.189e-02]
anha:  86%|########6 | 2845/3300 [30:11<12:50,  1.69s/it, lr: 1.0e-04 loss: 1.435e-01]
anha:  86%|########6 | 2845/3300 [30:11<12:50,  1.69s/it, lr: 1.0e-04 loss: 1.435e-01]
anha:  86%|########6 | 2846/3300 [30:11<16:09,  2.13s/it, lr: 1.0e-04 loss: 1.435e-01]
anha:  86%|########6 | 2846/3300 [30:11<16:09,  2.13s/it, lr: 1.0e-04 loss: 1.435e-01]
anha:  86%|########6 | 2846/3300 [30:13<16:09,  2.13s/it, lr: 1.0e-04 loss: 1.384e-01]
anha:  86%|########6 | 2846/3300 [30:13<16:09,  2.13s/it, lr: 1.0e-04 loss: 1.384e-01]
anha:  86%|########6 | 2847/3300 [30:13<15:13,  2.02s/it, lr: 1.0e-04 loss: 1.384e-01]
anha:  86%|########6 | 2847/3300 [30:13<15:13,  2.02s/it, lr: 1.0e-04 loss: 1.384e-01]
anha:  86%|########6 | 2847/3300 [30:14<15:13,  2.02s/it, lr: 1.0e-04 loss: 1.063e-01]
anha:  86%|########6 | 2847/3300 [30:14<15:13,  2.02s/it, lr: 1.0e-04 loss: 1.063e-01]
anha:  86%|########6 | 2848/3300 [30:14<15:03,  2.00s/it, lr: 1.0e-04 loss: 1.063e-01]
anha:  86%|########6 | 2848/3300 [30:14<15:03,  2.00s/it, lr: 1.0e-04 loss: 1.063e-01]
anha:  86%|########6 | 2848/3300 [30:15<15:03,  2.00s/it, lr: 1.0e-04 loss: 7.677e-02]
anha:  86%|########6 | 2848/3300 [30:15<15:03,  2.00s/it, lr: 1.0e-04 loss: 7.677e-02]
anha:  86%|########6 | 2849/3300 [30:15<12:47,  1.70s/it, lr: 1.0e-04 loss: 7.677e-02]
anha:  86%|########6 | 2849/3300 [30:15<12:47,  1.70s/it, lr: 1.0e-04 loss: 7.677e-02]
anha:  86%|########6 | 2849/3300 [30:19<12:47,  1.70s/it, lr: 1.0e-04 loss: 2.561e-01]
anha:  86%|########6 | 2849/3300 [30:19<12:47,  1.70s/it, lr: 1.0e-04 loss: 2.561e-01]
anha:  86%|########6 | 2850/3300 [30:21<12:46,  1.70s/it, lr: 1.0e-04 loss: 9.914e-02]
anha:  86%|########6 | 2850/3300 [30:21<12:46,  1.70s/it, lr: 1.0e-04 loss: 9.914e-02]
anha:  86%|########6 | 2851/3300 [30:21<11:12,  1.50s/it, lr: 1.0e-04 loss: 9.914e-02]
anha:  86%|########6 | 2851/3300 [30:21<11:12,  1.50s/it, lr: 1.0e-04 loss: 9.914e-02]
anha:  86%|########6 | 2851/3300 [30:23<11:12,  1.50s/it, lr: 1.0e-04 loss: 4.479e-02]
anha:  86%|########6 | 2851/3300 [30:23<11:12,  1.50s/it, lr: 1.0e-04 loss: 4.479e-02]
anha:  86%|########6 | 2852/3300 [30:23<12:05,  1.62s/it, lr: 1.0e-04 loss: 4.479e-02]
anha:  86%|########6 | 2852/3300 [30:23<12:05,  1.62s/it, lr: 1.0e-04 loss: 4.479e-02]
anha:  86%|########6 | 2852/3300 [30:25<12:05,  1.62s/it, lr: 1.0e-04 loss: 4.010e-02]
anha:  86%|########6 | 2852/3300 [30:25<12:05,  1.62s/it, lr: 1.0e-04 loss: 4.010e-02]
anha:  86%|########6 | 2853/3300 [30:25<12:15,  1.64s/it, lr: 1.0e-04 loss: 4.010e-02]
anha:  86%|########6 | 2853/3300 [30:25<12:15,  1.64s/it, lr: 1.0e-04 loss: 4.010e-02]
anha:  86%|########6 | 2853/3300 [30:27<12:15,  1.64s/it, lr: 1.0e-04 loss: 3.901e-02]
anha:  86%|########6 | 2853/3300 [30:27<12:15,  1.64s/it, lr: 1.0e-04 loss: 3.901e-02]
anha:  86%|########6 | 2854/3300 [30:27<13:01,  1.75s/it, lr: 1.0e-04 loss: 3.901e-02]
anha:  86%|########6 | 2854/3300 [30:27<13:01,  1.75s/it, lr: 1.0e-04 loss: 3.901e-02]
anha:  86%|########6 | 2854/3300 [30:28<13:01,  1.75s/it, lr: 1.0e-04 loss: 4.389e-02]
anha:  86%|########6 | 2854/3300 [30:28<13:01,  1.75s/it, lr: 1.0e-04 loss: 4.389e-02]
anha:  87%|########6 | 2855/3300 [30:28<11:29,  1.55s/it, lr: 1.0e-04 loss: 4.389e-02]
anha:  87%|########6 | 2855/3300 [30:28<11:29,  1.55s/it, lr: 1.0e-04 loss: 4.389e-02]
anha:  87%|########6 | 2855/3300 [30:29<11:29,  1.55s/it, lr: 1.0e-04 loss: 7.658e-02]
anha:  87%|########6 | 2855/3300 [30:29<11:29,  1.55s/it, lr: 1.0e-04 loss: 7.658e-02]
anha:  87%|########6 | 2856/3300 [30:29<10:17,  1.39s/it, lr: 1.0e-04 loss: 7.658e-02]
anha:  87%|########6 | 2856/3300 [30:29<10:17,  1.39s/it, lr: 1.0e-04 loss: 7.658e-02]
anha:  87%|########6 | 2856/3300 [30:33<10:17,  1.39s/it, lr: 1.0e-04 loss: 3.416e-02]
anha:  87%|########6 | 2856/3300 [30:33<10:17,  1.39s/it, lr: 1.0e-04 loss: 3.416e-02]
anha:  87%|########6 | 2857/3300 [30:33<14:43,  2.00s/it, lr: 1.0e-04 loss: 3.416e-02]
anha:  87%|########6 | 2857/3300 [30:33<14:43,  2.00s/it, lr: 1.0e-04 loss: 3.416e-02]
anha:  87%|########6 | 2857/3300 [30:34<14:43,  2.00s/it, lr: 1.0e-04 loss: 6.704e-02]
anha:  87%|########6 | 2857/3300 [30:34<14:43,  2.00s/it, lr: 1.0e-04 loss: 6.704e-02]
anha:  87%|########6 | 2858/3300 [30:34<12:34,  1.71s/it, lr: 1.0e-04 loss: 6.704e-02]
anha:  87%|########6 | 2858/3300 [30:34<12:34,  1.71s/it, lr: 1.0e-04 loss: 6.704e-02]
anha:  87%|########6 | 2858/3300 [30:37<12:34,  1.71s/it, lr: 1.0e-04 loss: 1.454e-01]
anha:  87%|########6 | 2858/3300 [30:37<12:34,  1.71s/it, lr: 1.0e-04 loss: 1.454e-01]
anha:  87%|########6 | 2859/3300 [30:37<16:08,  2.20s/it, lr: 1.0e-04 loss: 1.454e-01]
anha:  87%|########6 | 2859/3300 [30:37<16:08,  2.20s/it, lr: 1.0e-04 loss: 1.454e-01]
anha:  87%|########6 | 2859/3300 [30:38<16:08,  2.20s/it, lr: 1.0e-04 loss: 3.571e-02]
anha:  87%|########6 | 2859/3300 [30:38<16:08,  2.20s/it, lr: 1.0e-04 loss: 3.571e-02]
anha:  87%|########6 | 2860/3300 [30:41<16:06,  2.20s/it, lr: 1.0e-04 loss: 5.681e-02]
anha:  87%|########6 | 2860/3300 [30:41<16:06,  2.20s/it, lr: 1.0e-04 loss: 5.681e-02]
anha:  87%|########6 | 2861/3300 [30:41<14:13,  1.94s/it, lr: 1.0e-04 loss: 5.681e-02]
anha:  87%|########6 | 2861/3300 [30:41<14:13,  1.94s/it, lr: 1.0e-04 loss: 5.681e-02]
anha:  87%|########6 | 2861/3300 [30:45<14:13,  1.94s/it, lr: 1.0e-04 loss: 4.657e-02]
anha:  87%|########6 | 2861/3300 [30:45<14:13,  1.94s/it, lr: 1.0e-04 loss: 4.657e-02]
anha:  87%|########6 | 2862/3300 [30:45<16:35,  2.27s/it, lr: 1.0e-04 loss: 4.657e-02]
anha:  87%|########6 | 2862/3300 [30:45<16:35,  2.27s/it, lr: 1.0e-04 loss: 4.657e-02]
anha:  87%|########6 | 2862/3300 [30:47<16:35,  2.27s/it, lr: 1.0e-04 loss: 3.131e-02]
anha:  87%|########6 | 2862/3300 [30:47<16:35,  2.27s/it, lr: 1.0e-04 loss: 3.131e-02]
anha:  87%|########6 | 2863/3300 [30:47<16:11,  2.22s/it, lr: 1.0e-04 loss: 3.131e-02]
anha:  87%|########6 | 2863/3300 [30:47<16:11,  2.22s/it, lr: 1.0e-04 loss: 3.131e-02]
anha:  87%|########6 | 2863/3300 [30:50<16:11,  2.22s/it, lr: 1.0e-04 loss: 6.750e-02]
anha:  87%|########6 | 2863/3300 [30:50<16:11,  2.22s/it, lr: 1.0e-04 loss: 6.750e-02]
anha:  87%|########6 | 2864/3300 [30:50<18:24,  2.53s/it, lr: 1.0e-04 loss: 6.750e-02]
anha:  87%|########6 | 2864/3300 [30:50<18:24,  2.53s/it, lr: 1.0e-04 loss: 6.750e-02]
anha:  87%|########6 | 2864/3300 [30:51<18:24,  2.53s/it, lr: 1.0e-04 loss: 4.439e-02]
anha:  87%|########6 | 2864/3300 [30:51<18:24,  2.53s/it, lr: 1.0e-04 loss: 4.439e-02]
anha:  87%|########6 | 2865/3300 [30:51<15:19,  2.11s/it, lr: 1.0e-04 loss: 4.439e-02]
anha:  87%|########6 | 2865/3300 [30:51<15:19,  2.11s/it, lr: 1.0e-04 loss: 4.439e-02]
anha:  87%|########6 | 2865/3300 [30:52<15:19,  2.11s/it, lr: 1.0e-04 loss: 4.934e-02]
anha:  87%|########6 | 2865/3300 [30:52<15:19,  2.11s/it, lr: 1.0e-04 loss: 4.934e-02]
anha:  87%|########6 | 2866/3300 [30:52<12:59,  1.80s/it, lr: 1.0e-04 loss: 4.934e-02]
anha:  87%|########6 | 2866/3300 [30:52<12:59,  1.80s/it, lr: 1.0e-04 loss: 4.934e-02]
anha:  87%|########6 | 2866/3300 [30:54<12:59,  1.80s/it, lr: 1.0e-04 loss: 1.558e-01]
anha:  87%|########6 | 2866/3300 [30:54<12:59,  1.80s/it, lr: 1.0e-04 loss: 1.558e-01]
anha:  87%|########6 | 2867/3300 [30:54<13:19,  1.85s/it, lr: 1.0e-04 loss: 1.558e-01]
anha:  87%|########6 | 2867/3300 [30:54<13:19,  1.85s/it, lr: 1.0e-04 loss: 1.558e-01]
anha:  87%|########6 | 2867/3300 [30:56<13:19,  1.85s/it, lr: 1.0e-04 loss: 4.399e-02]
anha:  87%|########6 | 2867/3300 [30:56<13:19,  1.85s/it, lr: 1.0e-04 loss: 4.399e-02]
anha:  87%|########6 | 2868/3300 [30:56<13:34,  1.88s/it, lr: 1.0e-04 loss: 4.399e-02]
anha:  87%|########6 | 2868/3300 [30:56<13:34,  1.88s/it, lr: 1.0e-04 loss: 4.399e-02]
anha:  87%|########6 | 2868/3300 [30:57<13:34,  1.88s/it, lr: 1.0e-04 loss: 1.184e-01]
anha:  87%|########6 | 2868/3300 [30:57<13:34,  1.88s/it, lr: 1.0e-04 loss: 1.184e-01]
anha:  87%|########6 | 2869/3300 [30:57<11:39,  1.62s/it, lr: 1.0e-04 loss: 1.184e-01]
anha:  87%|########6 | 2869/3300 [30:57<11:39,  1.62s/it, lr: 1.0e-04 loss: 1.184e-01]
anha:  87%|########6 | 2869/3300 [31:00<11:39,  1.62s/it, lr: 1.0e-04 loss: 6.849e-02]
anha:  87%|########6 | 2869/3300 [31:00<11:39,  1.62s/it, lr: 1.0e-04 loss: 6.849e-02]
anha:  87%|########6 | 2870/3300 [31:04<11:37,  1.62s/it, lr: 1.0e-04 loss: 2.199e-01]
anha:  87%|########6 | 2870/3300 [31:04<11:37,  1.62s/it, lr: 1.0e-04 loss: 2.199e-01]
anha:  87%|########7 | 2871/3300 [31:04<11:59,  1.68s/it, lr: 1.0e-04 loss: 2.199e-01]
anha:  87%|########7 | 2871/3300 [31:04<11:59,  1.68s/it, lr: 1.0e-04 loss: 2.199e-01]
anha:  87%|########7 | 2871/3300 [31:07<11:59,  1.68s/it, lr: 1.0e-04 loss: 3.827e-02]
anha:  87%|########7 | 2871/3300 [31:07<11:59,  1.68s/it, lr: 1.0e-04 loss: 3.827e-02]
anha:  87%|########7 | 2872/3300 [31:07<15:00,  2.10s/it, lr: 1.0e-04 loss: 3.827e-02]
anha:  87%|########7 | 2872/3300 [31:07<15:00,  2.10s/it, lr: 1.0e-04 loss: 3.827e-02]
anha:  87%|########7 | 2872/3300 [31:11<15:00,  2.10s/it, lr: 1.0e-04 loss: 4.328e-02]
anha:  87%|########7 | 2872/3300 [31:11<15:00,  2.10s/it, lr: 1.0e-04 loss: 4.328e-02]
anha:  87%|########7 | 2873/3300 [31:11<17:31,  2.46s/it, lr: 1.0e-04 loss: 4.328e-02]
anha:  87%|########7 | 2873/3300 [31:11<17:31,  2.46s/it, lr: 1.0e-04 loss: 4.328e-02]
anha:  87%|########7 | 2873/3300 [31:13<17:31,  2.46s/it, lr: 1.0e-04 loss: 8.335e-02]
anha:  87%|########7 | 2873/3300 [31:13<17:31,  2.46s/it, lr: 1.0e-04 loss: 8.335e-02]
anha:  87%|########7 | 2874/3300 [31:13<16:33,  2.33s/it, lr: 1.0e-04 loss: 8.335e-02]
anha:  87%|########7 | 2874/3300 [31:13<16:33,  2.33s/it, lr: 1.0e-04 loss: 8.335e-02]
anha:  87%|########7 | 2874/3300 [31:14<16:33,  2.33s/it, lr: 1.0e-04 loss: 2.255e-01]
anha:  87%|########7 | 2874/3300 [31:14<16:33,  2.33s/it, lr: 1.0e-04 loss: 2.255e-01]
anha:  87%|########7 | 2875/3300 [31:14<13:53,  1.96s/it, lr: 1.0e-04 loss: 2.255e-01]
anha:  87%|########7 | 2875/3300 [31:14<13:53,  1.96s/it, lr: 1.0e-04 loss: 2.255e-01]
anha:  87%|########7 | 2875/3300 [31:17<13:53,  1.96s/it, lr: 1.0e-04 loss: 7.142e-02]
anha:  87%|########7 | 2875/3300 [31:17<13:53,  1.96s/it, lr: 1.0e-04 loss: 7.142e-02]
anha:  87%|########7 | 2876/3300 [31:17<16:42,  2.36s/it, lr: 1.0e-04 loss: 7.142e-02]
anha:  87%|########7 | 2876/3300 [31:17<16:42,  2.36s/it, lr: 1.0e-04 loss: 7.142e-02]
anha:  87%|########7 | 2876/3300 [31:18<16:42,  2.36s/it, lr: 1.0e-04 loss: 4.963e-02]
anha:  87%|########7 | 2876/3300 [31:18<16:42,  2.36s/it, lr: 1.0e-04 loss: 4.963e-02]
anha:  87%|########7 | 2877/3300 [31:18<13:53,  1.97s/it, lr: 1.0e-04 loss: 4.963e-02]
anha:  87%|########7 | 2877/3300 [31:18<13:53,  1.97s/it, lr: 1.0e-04 loss: 4.963e-02]
anha:  87%|########7 | 2877/3300 [31:19<13:53,  1.97s/it, lr: 1.0e-04 loss: 4.228e-02]
anha:  87%|########7 | 2877/3300 [31:19<13:53,  1.97s/it, lr: 1.0e-04 loss: 4.228e-02]
anha:  87%|########7 | 2878/3300 [31:19<11:50,  1.68s/it, lr: 1.0e-04 loss: 4.228e-02]
anha:  87%|########7 | 2878/3300 [31:19<11:50,  1.68s/it, lr: 1.0e-04 loss: 4.228e-02]
anha:  87%|########7 | 2878/3300 [31:21<11:50,  1.68s/it, lr: 1.0e-04 loss: 9.700e-02]
anha:  87%|########7 | 2878/3300 [31:21<11:50,  1.68s/it, lr: 1.0e-04 loss: 9.700e-02]
anha:  87%|########7 | 2879/3300 [31:21<12:22,  1.76s/it, lr: 1.0e-04 loss: 9.700e-02]
anha:  87%|########7 | 2879/3300 [31:21<12:22,  1.76s/it, lr: 1.0e-04 loss: 9.700e-02]
anha:  87%|########7 | 2879/3300 [31:23<12:22,  1.76s/it, lr: 1.0e-04 loss: 4.184e-02]
anha:  87%|########7 | 2879/3300 [31:23<12:22,  1.76s/it, lr: 1.0e-04 loss: 4.184e-02]
anha:  87%|########7 | 2880/3300 [31:24<12:20,  1.76s/it, lr: 1.0e-04 loss: 4.160e-02]
anha:  87%|########7 | 2880/3300 [31:24<12:20,  1.76s/it, lr: 1.0e-04 loss: 4.160e-02]
anha:  87%|########7 | 2881/3300 [31:24<08:17,  1.19s/it, lr: 1.0e-04 loss: 4.160e-02]
anha:  87%|########7 | 2881/3300 [31:24<08:17,  1.19s/it, lr: 1.0e-04 loss: 4.160e-02]
anha:  87%|########7 | 2881/3300 [31:25<08:17,  1.19s/it, lr: 1.0e-04 loss: 4.263e-02]
anha:  87%|########7 | 2881/3300 [31:25<08:17,  1.19s/it, lr: 1.0e-04 loss: 4.263e-02]
anha:  87%|########7 | 2882/3300 [31:25<08:07,  1.17s/it, lr: 1.0e-04 loss: 4.263e-02]
anha:  87%|########7 | 2882/3300 [31:25<08:07,  1.17s/it, lr: 1.0e-04 loss: 4.263e-02]
anha:  87%|########7 | 2882/3300 [31:26<08:07,  1.17s/it, lr: 1.0e-04 loss: 6.620e-02]
anha:  87%|########7 | 2882/3300 [31:26<08:07,  1.17s/it, lr: 1.0e-04 loss: 6.620e-02]
anha:  87%|########7 | 2883/3300 [31:26<07:50,  1.13s/it, lr: 1.0e-04 loss: 6.620e-02]
anha:  87%|########7 | 2883/3300 [31:26<07:50,  1.13s/it, lr: 1.0e-04 loss: 6.620e-02]
anha:  87%|########7 | 2883/3300 [31:29<07:50,  1.13s/it, lr: 1.0e-04 loss: 4.390e-02]
anha:  87%|########7 | 2883/3300 [31:29<07:50,  1.13s/it, lr: 1.0e-04 loss: 4.390e-02]
anha:  87%|########7 | 2884/3300 [31:29<11:38,  1.68s/it, lr: 1.0e-04 loss: 4.390e-02]
anha:  87%|########7 | 2884/3300 [31:29<11:38,  1.68s/it, lr: 1.0e-04 loss: 4.390e-02]
anha:  87%|########7 | 2884/3300 [31:30<11:38,  1.68s/it, lr: 1.0e-04 loss: 6.253e-02]
anha:  87%|########7 | 2884/3300 [31:30<11:38,  1.68s/it, lr: 1.0e-04 loss: 6.253e-02]
anha:  87%|########7 | 2885/3300 [31:30<10:19,  1.49s/it, lr: 1.0e-04 loss: 6.253e-02]
anha:  87%|########7 | 2885/3300 [31:30<10:19,  1.49s/it, lr: 1.0e-04 loss: 6.253e-02]
anha:  87%|########7 | 2885/3300 [31:34<10:19,  1.49s/it, lr: 1.0e-04 loss: 4.642e-02]
anha:  87%|########7 | 2885/3300 [31:34<10:19,  1.49s/it, lr: 1.0e-04 loss: 4.642e-02]
anha:  87%|########7 | 2886/3300 [31:34<14:11,  2.06s/it, lr: 1.0e-04 loss: 4.642e-02]
anha:  87%|########7 | 2886/3300 [31:34<14:11,  2.06s/it, lr: 1.0e-04 loss: 4.642e-02]
anha:  87%|########7 | 2886/3300 [31:36<14:11,  2.06s/it, lr: 1.0e-04 loss: 5.929e-02]
anha:  87%|########7 | 2886/3300 [31:36<14:11,  2.06s/it, lr: 1.0e-04 loss: 5.929e-02]
anha:  87%|########7 | 2887/3300 [31:36<13:30,  1.96s/it, lr: 1.0e-04 loss: 5.929e-02]
anha:  87%|########7 | 2887/3300 [31:36<13:30,  1.96s/it, lr: 1.0e-04 loss: 5.929e-02]
anha:  87%|########7 | 2887/3300 [31:38<13:30,  1.96s/it, lr: 1.0e-04 loss: 5.674e-02]
anha:  87%|########7 | 2887/3300 [31:38<13:30,  1.96s/it, lr: 1.0e-04 loss: 5.674e-02]
anha:  88%|########7 | 2888/3300 [31:38<13:27,  1.96s/it, lr: 1.0e-04 loss: 5.674e-02]
anha:  88%|########7 | 2888/3300 [31:38<13:27,  1.96s/it, lr: 1.0e-04 loss: 5.674e-02]
anha:  88%|########7 | 2888/3300 [31:39<13:27,  1.96s/it, lr: 1.0e-04 loss: 4.664e-02]
anha:  88%|########7 | 2888/3300 [31:39<13:27,  1.96s/it, lr: 1.0e-04 loss: 4.664e-02]
anha:  88%|########7 | 2889/3300 [31:39<13:24,  1.96s/it, lr: 1.0e-04 loss: 4.664e-02]
anha:  88%|########7 | 2889/3300 [31:39<13:24,  1.96s/it, lr: 1.0e-04 loss: 4.664e-02]
anha:  88%|########7 | 2889/3300 [31:41<13:24,  1.96s/it, lr: 1.0e-04 loss: 6.047e-02]
anha:  88%|########7 | 2889/3300 [31:41<13:24,  1.96s/it, lr: 1.0e-04 loss: 6.047e-02]
anha:  88%|########7 | 2890/3300 [31:45<13:22,  1.96s/it, lr: 1.0e-04 loss: 4.677e-02]
anha:  88%|########7 | 2890/3300 [31:45<13:22,  1.96s/it, lr: 1.0e-04 loss: 4.677e-02]
anha:  88%|########7 | 2891/3300 [31:45<12:24,  1.82s/it, lr: 1.0e-04 loss: 4.677e-02]
anha:  88%|########7 | 2891/3300 [31:45<12:24,  1.82s/it, lr: 1.0e-04 loss: 4.677e-02]
anha:  88%|########7 | 2891/3300 [31:47<12:24,  1.82s/it, lr: 1.0e-04 loss: 3.529e-02]
anha:  88%|########7 | 2891/3300 [31:47<12:24,  1.82s/it, lr: 1.0e-04 loss: 3.529e-02]
anha:  88%|########7 | 2892/3300 [31:47<12:38,  1.86s/it, lr: 1.0e-04 loss: 3.529e-02]
anha:  88%|########7 | 2892/3300 [31:47<12:38,  1.86s/it, lr: 1.0e-04 loss: 3.529e-02]
anha:  88%|########7 | 2892/3300 [31:48<12:38,  1.86s/it, lr: 1.0e-04 loss: 6.584e-02]
anha:  88%|########7 | 2892/3300 [31:48<12:38,  1.86s/it, lr: 1.0e-04 loss: 6.584e-02]
anha:  88%|########7 | 2893/3300 [31:48<11:05,  1.63s/it, lr: 1.0e-04 loss: 6.584e-02]
anha:  88%|########7 | 2893/3300 [31:48<11:05,  1.63s/it, lr: 1.0e-04 loss: 6.584e-02]
anha:  88%|########7 | 2893/3300 [31:51<11:05,  1.63s/it, lr: 1.0e-04 loss: 7.731e-02]
anha:  88%|########7 | 2893/3300 [31:51<11:05,  1.63s/it, lr: 1.0e-04 loss: 7.731e-02]
anha:  88%|########7 | 2894/3300 [31:51<13:59,  2.07s/it, lr: 1.0e-04 loss: 7.731e-02]
anha:  88%|########7 | 2894/3300 [31:51<13:59,  2.07s/it, lr: 1.0e-04 loss: 7.731e-02]
anha:  88%|########7 | 2894/3300 [31:53<13:59,  2.07s/it, lr: 1.0e-04 loss: 4.918e-02]
anha:  88%|########7 | 2894/3300 [31:53<13:59,  2.07s/it, lr: 1.0e-04 loss: 4.918e-02]
anha:  88%|########7 | 2895/3300 [31:53<13:19,  1.98s/it, lr: 1.0e-04 loss: 4.918e-02]
anha:  88%|########7 | 2895/3300 [31:53<13:19,  1.98s/it, lr: 1.0e-04 loss: 4.918e-02]
anha:  88%|########7 | 2895/3300 [31:54<13:19,  1.98s/it, lr: 1.0e-04 loss: 5.473e-02]
anha:  88%|########7 | 2895/3300 [31:54<13:19,  1.98s/it, lr: 1.0e-04 loss: 5.473e-02]
anha:  88%|########7 | 2896/3300 [31:54<11:26,  1.70s/it, lr: 1.0e-04 loss: 5.473e-02]
anha:  88%|########7 | 2896/3300 [31:54<11:26,  1.70s/it, lr: 1.0e-04 loss: 5.473e-02]
anha:  88%|########7 | 2896/3300 [31:55<11:26,  1.70s/it, lr: 1.0e-04 loss: 9.667e-02]
anha:  88%|########7 | 2896/3300 [31:55<11:26,  1.70s/it, lr: 1.0e-04 loss: 9.667e-02]
anha:  88%|########7 | 2897/3300 [31:55<10:01,  1.49s/it, lr: 1.0e-04 loss: 9.667e-02]
anha:  88%|########7 | 2897/3300 [31:55<10:01,  1.49s/it, lr: 1.0e-04 loss: 9.667e-02]
anha:  88%|########7 | 2897/3300 [31:56<10:01,  1.49s/it, lr: 1.0e-04 loss: 7.060e-02]
anha:  88%|########7 | 2897/3300 [31:56<10:01,  1.49s/it, lr: 1.0e-04 loss: 7.060e-02]
anha:  88%|########7 | 2898/3300 [31:56<10:29,  1.57s/it, lr: 1.0e-04 loss: 7.060e-02]
anha:  88%|########7 | 2898/3300 [31:56<10:29,  1.57s/it, lr: 1.0e-04 loss: 7.060e-02]
anha:  88%|########7 | 2898/3300 [32:00<10:29,  1.57s/it, lr: 1.0e-04 loss: 2.844e-02]
anha:  88%|########7 | 2898/3300 [32:00<10:29,  1.57s/it, lr: 1.0e-04 loss: 2.844e-02]
anha:  88%|########7 | 2899/3300 [32:00<14:01,  2.10s/it, lr: 1.0e-04 loss: 2.844e-02]
anha:  88%|########7 | 2899/3300 [32:00<14:01,  2.10s/it, lr: 1.0e-04 loss: 2.844e-02]
anha:  88%|########7 | 2899/3300 [32:02<14:01,  2.10s/it, lr: 1.0e-04 loss: 4.154e-02]
anha:  88%|########7 | 2899/3300 [32:02<14:01,  2.10s/it, lr: 1.0e-04 loss: 4.154e-02]
anha:  88%|########7 | 2900/3300 [32:05<13:59,  2.10s/it, lr: 1.0e-04 loss: 1.067e-01]
anha:  88%|########7 | 2900/3300 [32:05<13:59,  2.10s/it, lr: 1.0e-04 loss: 1.067e-01]
anha:  88%|########7 | 2901/3300 [32:05<12:51,  1.93s/it, lr: 1.0e-04 loss: 1.067e-01]
anha:  88%|########7 | 2901/3300 [32:05<12:51,  1.93s/it, lr: 1.0e-04 loss: 1.067e-01]
anha:  88%|########7 | 2901/3300 [32:08<12:51,  1.93s/it, lr: 1.0e-04 loss: 6.535e-02]
anha:  88%|########7 | 2901/3300 [32:08<12:51,  1.93s/it, lr: 1.0e-04 loss: 6.535e-02]
anha:  88%|########7 | 2902/3300 [32:08<15:09,  2.29s/it, lr: 1.0e-04 loss: 6.535e-02]
anha:  88%|########7 | 2902/3300 [32:08<15:09,  2.29s/it, lr: 1.0e-04 loss: 6.535e-02]
anha:  88%|########7 | 2902/3300 [32:09<15:09,  2.29s/it, lr: 1.0e-04 loss: 6.081e-02]
anha:  88%|########7 | 2902/3300 [32:09<15:09,  2.29s/it, lr: 1.0e-04 loss: 6.081e-02]
anha:  88%|########7 | 2903/3300 [32:09<12:57,  1.96s/it, lr: 1.0e-04 loss: 6.081e-02]
anha:  88%|########7 | 2903/3300 [32:09<12:57,  1.96s/it, lr: 1.0e-04 loss: 6.081e-02]
anha:  88%|########7 | 2903/3300 [32:11<12:57,  1.96s/it, lr: 1.0e-04 loss: 5.857e-02]
anha:  88%|########7 | 2903/3300 [32:11<12:57,  1.96s/it, lr: 1.0e-04 loss: 5.857e-02]
anha:  88%|########8 | 2904/3300 [32:11<11:13,  1.70s/it, lr: 1.0e-04 loss: 5.857e-02]
anha:  88%|########8 | 2904/3300 [32:11<11:13,  1.70s/it, lr: 1.0e-04 loss: 5.857e-02]
anha:  88%|########8 | 2904/3300 [32:14<11:13,  1.70s/it, lr: 1.0e-04 loss: 5.702e-02]
anha:  88%|########8 | 2904/3300 [32:14<11:13,  1.70s/it, lr: 1.0e-04 loss: 5.702e-02]
anha:  88%|########8 | 2905/3300 [32:14<14:16,  2.17s/it, lr: 1.0e-04 loss: 5.702e-02]
anha:  88%|########8 | 2905/3300 [32:14<14:16,  2.17s/it, lr: 1.0e-04 loss: 5.702e-02]
anha:  88%|########8 | 2905/3300 [32:17<14:16,  2.17s/it, lr: 1.0e-04 loss: 7.370e-02]
anha:  88%|########8 | 2905/3300 [32:17<14:16,  2.17s/it, lr: 1.0e-04 loss: 7.370e-02]
anha:  88%|########8 | 2906/3300 [32:17<16:14,  2.47s/it, lr: 1.0e-04 loss: 7.370e-02]
anha:  88%|########8 | 2906/3300 [32:17<16:14,  2.47s/it, lr: 1.0e-04 loss: 7.370e-02]
anha:  88%|########8 | 2906/3300 [32:21<16:14,  2.47s/it, lr: 1.0e-04 loss: 4.598e-02]
anha:  88%|########8 | 2906/3300 [32:21<16:14,  2.47s/it, lr: 1.0e-04 loss: 4.598e-02]
anha:  88%|########8 | 2907/3300 [32:21<18:06,  2.76s/it, lr: 1.0e-04 loss: 4.598e-02]
anha:  88%|########8 | 2907/3300 [32:21<18:06,  2.76s/it, lr: 1.0e-04 loss: 4.598e-02]
anha:  88%|########8 | 2907/3300 [32:22<18:06,  2.76s/it, lr: 1.0e-04 loss: 3.562e-02]
anha:  88%|########8 | 2907/3300 [32:22<18:06,  2.76s/it, lr: 1.0e-04 loss: 3.562e-02]
anha:  88%|########8 | 2908/3300 [32:22<14:44,  2.26s/it, lr: 1.0e-04 loss: 3.562e-02]
anha:  88%|########8 | 2908/3300 [32:22<14:44,  2.26s/it, lr: 1.0e-04 loss: 3.562e-02]
anha:  88%|########8 | 2908/3300 [32:23<14:44,  2.26s/it, lr: 1.0e-04 loss: 6.091e-02]
anha:  88%|########8 | 2908/3300 [32:23<14:44,  2.26s/it, lr: 1.0e-04 loss: 6.091e-02]
anha:  88%|########8 | 2909/3300 [32:23<12:17,  1.89s/it, lr: 1.0e-04 loss: 6.091e-02]
anha:  88%|########8 | 2909/3300 [32:23<12:17,  1.89s/it, lr: 1.0e-04 loss: 6.091e-02]
anha:  88%|########8 | 2909/3300 [32:26<12:17,  1.89s/it, lr: 1.0e-04 loss: 3.604e-02]
anha:  88%|########8 | 2909/3300 [32:26<12:17,  1.89s/it, lr: 1.0e-04 loss: 3.604e-02]
anha:  88%|########8 | 2910/3300 [32:28<12:15,  1.89s/it, lr: 1.0e-04 loss: 3.431e-02]
anha:  88%|########8 | 2910/3300 [32:28<12:15,  1.89s/it, lr: 1.0e-04 loss: 3.431e-02]
anha:  88%|########8 | 2911/3300 [32:28<09:31,  1.47s/it, lr: 1.0e-04 loss: 3.431e-02]
anha:  88%|########8 | 2911/3300 [32:28<09:31,  1.47s/it, lr: 1.0e-04 loss: 3.431e-02]
anha:  88%|########8 | 2911/3300 [32:29<09:31,  1.47s/it, lr: 1.0e-04 loss: 5.844e-02]
anha:  88%|########8 | 2911/3300 [32:29<09:31,  1.47s/it, lr: 1.0e-04 loss: 5.844e-02]
anha:  88%|########8 | 2912/3300 [32:29<08:45,  1.35s/it, lr: 1.0e-04 loss: 5.844e-02]
anha:  88%|########8 | 2912/3300 [32:29<08:45,  1.35s/it, lr: 1.0e-04 loss: 5.844e-02]
anha:  88%|########8 | 2912/3300 [32:31<08:45,  1.35s/it, lr: 1.0e-04 loss: 9.655e-02]
anha:  88%|########8 | 2912/3300 [32:31<08:45,  1.35s/it, lr: 1.0e-04 loss: 9.655e-02]
anha:  88%|########8 | 2913/3300 [32:31<09:47,  1.52s/it, lr: 1.0e-04 loss: 9.655e-02]
anha:  88%|########8 | 2913/3300 [32:31<09:47,  1.52s/it, lr: 1.0e-04 loss: 9.655e-02]
anha:  88%|########8 | 2913/3300 [32:34<09:47,  1.52s/it, lr: 1.0e-04 loss: 3.416e-02]
anha:  88%|########8 | 2913/3300 [32:34<09:47,  1.52s/it, lr: 1.0e-04 loss: 3.416e-02]
anha:  88%|########8 | 2914/3300 [32:34<12:59,  2.02s/it, lr: 1.0e-04 loss: 3.416e-02]
anha:  88%|########8 | 2914/3300 [32:34<12:59,  2.02s/it, lr: 1.0e-04 loss: 3.416e-02]
anha:  88%|########8 | 2914/3300 [32:38<12:59,  2.02s/it, lr: 1.0e-04 loss: 8.853e-02]
anha:  88%|########8 | 2914/3300 [32:38<12:59,  2.02s/it, lr: 1.0e-04 loss: 8.853e-02]
anha:  88%|########8 | 2915/3300 [32:38<15:36,  2.43s/it, lr: 1.0e-04 loss: 8.853e-02]
anha:  88%|########8 | 2915/3300 [32:38<15:36,  2.43s/it, lr: 1.0e-04 loss: 8.853e-02]
anha:  88%|########8 | 2915/3300 [32:40<15:36,  2.43s/it, lr: 1.0e-04 loss: 8.561e-02]
anha:  88%|########8 | 2915/3300 [32:40<15:36,  2.43s/it, lr: 1.0e-04 loss: 8.561e-02]
anha:  88%|########8 | 2916/3300 [32:40<14:18,  2.24s/it, lr: 1.0e-04 loss: 8.561e-02]
anha:  88%|########8 | 2916/3300 [32:40<14:18,  2.24s/it, lr: 1.0e-04 loss: 8.561e-02]
anha:  88%|########8 | 2916/3300 [32:43<14:18,  2.24s/it, lr: 1.0e-04 loss: 5.277e-02]
anha:  88%|########8 | 2916/3300 [32:43<14:18,  2.24s/it, lr: 1.0e-04 loss: 5.277e-02]
anha:  88%|########8 | 2917/3300 [32:43<16:22,  2.57s/it, lr: 1.0e-04 loss: 5.277e-02]
anha:  88%|########8 | 2917/3300 [32:43<16:22,  2.57s/it, lr: 1.0e-04 loss: 5.277e-02]
anha:  88%|########8 | 2917/3300 [32:46<16:22,  2.57s/it, lr: 1.0e-04 loss: 1.816e-01]
anha:  88%|########8 | 2917/3300 [32:46<16:22,  2.57s/it, lr: 1.0e-04 loss: 1.816e-01]
anha:  88%|########8 | 2918/3300 [32:46<17:40,  2.78s/it, lr: 1.0e-04 loss: 1.816e-01]
anha:  88%|########8 | 2918/3300 [32:46<17:40,  2.78s/it, lr: 1.0e-04 loss: 1.816e-01]
anha:  88%|########8 | 2918/3300 [32:48<17:40,  2.78s/it, lr: 1.0e-04 loss: 3.247e-02]
anha:  88%|########8 | 2918/3300 [32:48<17:40,  2.78s/it, lr: 1.0e-04 loss: 3.247e-02]
anha:  88%|########8 | 2919/3300 [32:48<16:16,  2.56s/it, lr: 1.0e-04 loss: 3.247e-02]
anha:  88%|########8 | 2919/3300 [32:48<16:16,  2.56s/it, lr: 1.0e-04 loss: 3.247e-02]
anha:  88%|########8 | 2919/3300 [32:49<16:16,  2.56s/it, lr: 1.0e-04 loss: 8.440e-02]
anha:  88%|########8 | 2919/3300 [32:49<16:16,  2.56s/it, lr: 1.0e-04 loss: 8.440e-02]
anha:  88%|########8 | 2920/3300 [32:50<16:13,  2.56s/it, lr: 1.0e-04 loss: 3.854e-02]
anha:  88%|########8 | 2920/3300 [32:50<16:13,  2.56s/it, lr: 1.0e-04 loss: 3.854e-02]
anha:  89%|########8 | 2921/3300 [32:50<10:14,  1.62s/it, lr: 1.0e-04 loss: 3.854e-02]
anha:  89%|########8 | 2921/3300 [32:50<10:14,  1.62s/it, lr: 1.0e-04 loss: 3.854e-02]
anha:  89%|########8 | 2921/3300 [32:52<10:14,  1.62s/it, lr: 1.0e-04 loss: 6.086e-02]
anha:  89%|########8 | 2921/3300 [32:52<10:14,  1.62s/it, lr: 1.0e-04 loss: 6.086e-02]
anha:  89%|########8 | 2922/3300 [32:52<10:23,  1.65s/it, lr: 1.0e-04 loss: 6.086e-02]
anha:  89%|########8 | 2922/3300 [32:52<10:23,  1.65s/it, lr: 1.0e-04 loss: 6.086e-02]
anha:  89%|########8 | 2922/3300 [32:53<10:23,  1.65s/it, lr: 1.0e-04 loss: 4.787e-02]
anha:  89%|########8 | 2922/3300 [32:53<10:23,  1.65s/it, lr: 1.0e-04 loss: 4.787e-02]
anha:  89%|########8 | 2923/3300 [32:53<09:16,  1.48s/it, lr: 1.0e-04 loss: 4.787e-02]
anha:  89%|########8 | 2923/3300 [32:53<09:16,  1.48s/it, lr: 1.0e-04 loss: 4.787e-02]
anha:  89%|########8 | 2923/3300 [32:55<09:16,  1.48s/it, lr: 1.0e-04 loss: 2.481e-02]
anha:  89%|########8 | 2923/3300 [32:55<09:16,  1.48s/it, lr: 1.0e-04 loss: 2.481e-02]
anha:  89%|########8 | 2924/3300 [32:55<10:03,  1.61s/it, lr: 1.0e-04 loss: 2.481e-02]
anha:  89%|########8 | 2924/3300 [32:55<10:03,  1.61s/it, lr: 1.0e-04 loss: 2.481e-02]
anha:  89%|########8 | 2924/3300 [32:57<10:03,  1.61s/it, lr: 1.0e-04 loss: 7.067e-02]
anha:  89%|########8 | 2924/3300 [32:57<10:03,  1.61s/it, lr: 1.0e-04 loss: 7.067e-02]
anha:  89%|########8 | 2925/3300 [32:57<10:17,  1.65s/it, lr: 1.0e-04 loss: 7.067e-02]
anha:  89%|########8 | 2925/3300 [32:57<10:17,  1.65s/it, lr: 1.0e-04 loss: 7.067e-02]
anha:  89%|########8 | 2925/3300 [32:59<10:17,  1.65s/it, lr: 1.0e-04 loss: 9.884e-02]
anha:  89%|########8 | 2925/3300 [32:59<10:17,  1.65s/it, lr: 1.0e-04 loss: 9.884e-02]
anha:  89%|########8 | 2926/3300 [32:59<10:49,  1.74s/it, lr: 1.0e-04 loss: 9.884e-02]
anha:  89%|########8 | 2926/3300 [32:59<10:49,  1.74s/it, lr: 1.0e-04 loss: 9.884e-02]
anha:  89%|########8 | 2926/3300 [33:01<10:49,  1.74s/it, lr: 1.0e-04 loss: 6.261e-02]
anha:  89%|########8 | 2926/3300 [33:01<10:49,  1.74s/it, lr: 1.0e-04 loss: 6.261e-02]
anha:  89%|########8 | 2927/3300 [33:01<11:12,  1.80s/it, lr: 1.0e-04 loss: 6.261e-02]
anha:  89%|########8 | 2927/3300 [33:01<11:12,  1.80s/it, lr: 1.0e-04 loss: 6.261e-02]
anha:  89%|########8 | 2927/3300 [33:04<11:12,  1.80s/it, lr: 1.0e-04 loss: 4.333e-02]
anha:  89%|########8 | 2927/3300 [33:04<11:12,  1.80s/it, lr: 1.0e-04 loss: 4.333e-02]
anha:  89%|########8 | 2928/3300 [33:04<14:16,  2.30s/it, lr: 1.0e-04 loss: 4.333e-02]
anha:  89%|########8 | 2928/3300 [33:04<14:16,  2.30s/it, lr: 1.0e-04 loss: 4.333e-02]
anha:  89%|########8 | 2928/3300 [33:08<14:16,  2.30s/it, lr: 1.0e-04 loss: 4.196e-02]
anha:  89%|########8 | 2928/3300 [33:08<14:16,  2.30s/it, lr: 1.0e-04 loss: 4.196e-02]
anha:  89%|########8 | 2929/3300 [33:08<16:01,  2.59s/it, lr: 1.0e-04 loss: 4.196e-02]
anha:  89%|########8 | 2929/3300 [33:08<16:01,  2.59s/it, lr: 1.0e-04 loss: 4.196e-02]
anha:  89%|########8 | 2929/3300 [33:09<16:01,  2.59s/it, lr: 1.0e-04 loss: 3.440e-02]
anha:  89%|########8 | 2929/3300 [33:09<16:01,  2.59s/it, lr: 1.0e-04 loss: 3.440e-02]
anha:  89%|########8 | 2930/3300 [33:11<15:59,  2.59s/it, lr: 1.0e-04 loss: 5.913e-02]
anha:  89%|########8 | 2930/3300 [33:11<15:59,  2.59s/it, lr: 1.0e-04 loss: 5.913e-02]
anha:  89%|########8 | 2931/3300 [33:11<11:23,  1.85s/it, lr: 1.0e-04 loss: 5.913e-02]
anha:  89%|########8 | 2931/3300 [33:11<11:23,  1.85s/it, lr: 1.0e-04 loss: 5.913e-02]
anha:  89%|########8 | 2931/3300 [33:14<11:23,  1.85s/it, lr: 1.0e-04 loss: 1.446e-01]
anha:  89%|########8 | 2931/3300 [33:14<11:23,  1.85s/it, lr: 1.0e-04 loss: 1.446e-01]
anha:  89%|########8 | 2932/3300 [33:14<13:45,  2.24s/it, lr: 1.0e-04 loss: 1.446e-01]
anha:  89%|########8 | 2932/3300 [33:14<13:45,  2.24s/it, lr: 1.0e-04 loss: 1.446e-01]
anha:  89%|########8 | 2932/3300 [33:15<13:45,  2.24s/it, lr: 1.0e-04 loss: 1.467e-01]
anha:  89%|########8 | 2932/3300 [33:15<13:45,  2.24s/it, lr: 1.0e-04 loss: 1.467e-01]
anha:  89%|########8 | 2933/3300 [33:15<11:45,  1.92s/it, lr: 1.0e-04 loss: 1.467e-01]
anha:  89%|########8 | 2933/3300 [33:15<11:45,  1.92s/it, lr: 1.0e-04 loss: 1.467e-01]
anha:  89%|########8 | 2933/3300 [33:18<11:45,  1.92s/it, lr: 1.0e-04 loss: 7.096e-02]
anha:  89%|########8 | 2933/3300 [33:18<11:45,  1.92s/it, lr: 1.0e-04 loss: 7.096e-02]
anha:  89%|########8 | 2934/3300 [33:18<14:14,  2.33s/it, lr: 1.0e-04 loss: 7.096e-02]
anha:  89%|########8 | 2934/3300 [33:18<14:14,  2.33s/it, lr: 1.0e-04 loss: 7.096e-02]
anha:  89%|########8 | 2934/3300 [33:20<14:14,  2.33s/it, lr: 1.0e-04 loss: 3.313e-02]
anha:  89%|########8 | 2934/3300 [33:20<14:14,  2.33s/it, lr: 1.0e-04 loss: 3.313e-02]
anha:  89%|########8 | 2935/3300 [33:20<13:35,  2.23s/it, lr: 1.0e-04 loss: 3.313e-02]
anha:  89%|########8 | 2935/3300 [33:20<13:35,  2.23s/it, lr: 1.0e-04 loss: 3.313e-02]
anha:  89%|########8 | 2935/3300 [33:21<13:35,  2.23s/it, lr: 1.0e-04 loss: 6.568e-02]
anha:  89%|########8 | 2935/3300 [33:21<13:35,  2.23s/it, lr: 1.0e-04 loss: 6.568e-02]
anha:  89%|########8 | 2936/3300 [33:21<11:24,  1.88s/it, lr: 1.0e-04 loss: 6.568e-02]
anha:  89%|########8 | 2936/3300 [33:21<11:24,  1.88s/it, lr: 1.0e-04 loss: 6.568e-02]
anha:  89%|########8 | 2936/3300 [33:23<11:24,  1.88s/it, lr: 1.0e-04 loss: 5.599e-02]
anha:  89%|########8 | 2936/3300 [33:23<11:24,  1.88s/it, lr: 1.0e-04 loss: 5.599e-02]
anha:  89%|########9 | 2937/3300 [33:23<11:32,  1.91s/it, lr: 1.0e-04 loss: 5.599e-02]
anha:  89%|########9 | 2937/3300 [33:23<11:32,  1.91s/it, lr: 1.0e-04 loss: 5.599e-02]
anha:  89%|########9 | 2937/3300 [33:27<11:32,  1.91s/it, lr: 1.0e-04 loss: 4.017e-02]
anha:  89%|########9 | 2937/3300 [33:27<11:32,  1.91s/it, lr: 1.0e-04 loss: 4.017e-02]
anha:  89%|########9 | 2938/3300 [33:27<13:55,  2.31s/it, lr: 1.0e-04 loss: 4.017e-02]
anha:  89%|########9 | 2938/3300 [33:27<13:55,  2.31s/it, lr: 1.0e-04 loss: 4.017e-02]
anha:  89%|########9 | 2938/3300 [33:28<13:55,  2.31s/it, lr: 1.0e-04 loss: 8.709e-02]
anha:  89%|########9 | 2938/3300 [33:28<13:55,  2.31s/it, lr: 1.0e-04 loss: 8.709e-02]
anha:  89%|########9 | 2939/3300 [33:28<12:50,  2.13s/it, lr: 1.0e-04 loss: 8.709e-02]
anha:  89%|########9 | 2939/3300 [33:28<12:50,  2.13s/it, lr: 1.0e-04 loss: 8.709e-02]
anha:  89%|########9 | 2939/3300 [33:29<12:50,  2.13s/it, lr: 1.0e-04 loss: 8.727e-02]
anha:  89%|########9 | 2939/3300 [33:29<12:50,  2.13s/it, lr: 1.0e-04 loss: 8.727e-02]
anha:  89%|########9 | 2940/3300 [33:31<12:47,  2.13s/it, lr: 1.0e-04 loss: 4.432e-02]
anha:  89%|########9 | 2940/3300 [33:31<12:47,  2.13s/it, lr: 1.0e-04 loss: 4.432e-02]
anha:  89%|########9 | 2941/3300 [33:31<09:37,  1.61s/it, lr: 1.0e-04 loss: 4.432e-02]
anha:  89%|########9 | 2941/3300 [33:31<09:37,  1.61s/it, lr: 1.0e-04 loss: 4.432e-02]
anha:  89%|########9 | 2941/3300 [33:35<09:37,  1.61s/it, lr: 1.0e-04 loss: 3.614e-02]
anha:  89%|########9 | 2941/3300 [33:35<09:37,  1.61s/it, lr: 1.0e-04 loss: 3.614e-02]
anha:  89%|########9 | 2942/3300 [33:35<12:11,  2.04s/it, lr: 1.0e-04 loss: 3.614e-02]
anha:  89%|########9 | 2942/3300 [33:35<12:11,  2.04s/it, lr: 1.0e-04 loss: 3.614e-02]
anha:  89%|########9 | 2942/3300 [33:38<12:11,  2.04s/it, lr: 1.0e-04 loss: 9.074e-02]
anha:  89%|########9 | 2942/3300 [33:38<12:11,  2.04s/it, lr: 1.0e-04 loss: 9.074e-02]
anha:  89%|########9 | 2943/3300 [33:38<14:14,  2.39s/it, lr: 1.0e-04 loss: 9.074e-02]
anha:  89%|########9 | 2943/3300 [33:38<14:14,  2.39s/it, lr: 1.0e-04 loss: 9.074e-02]
anha:  89%|########9 | 2943/3300 [33:42<14:14,  2.39s/it, lr: 1.0e-04 loss: 1.007e-01]
anha:  89%|########9 | 2943/3300 [33:42<14:14,  2.39s/it, lr: 1.0e-04 loss: 1.007e-01]
anha:  89%|########9 | 2944/3300 [33:42<15:46,  2.66s/it, lr: 1.0e-04 loss: 1.007e-01]
anha:  89%|########9 | 2944/3300 [33:42<15:46,  2.66s/it, lr: 1.0e-04 loss: 1.007e-01]
anha:  89%|########9 | 2944/3300 [33:43<15:46,  2.66s/it, lr: 1.0e-04 loss: 4.017e-02]
anha:  89%|########9 | 2944/3300 [33:43<15:46,  2.66s/it, lr: 1.0e-04 loss: 4.017e-02]
anha:  89%|########9 | 2945/3300 [33:43<13:00,  2.20s/it, lr: 1.0e-04 loss: 4.017e-02]
anha:  89%|########9 | 2945/3300 [33:43<13:00,  2.20s/it, lr: 1.0e-04 loss: 4.017e-02]
anha:  89%|########9 | 2945/3300 [33:44<13:00,  2.20s/it, lr: 1.0e-04 loss: 5.746e-02]
anha:  89%|########9 | 2945/3300 [33:44<13:00,  2.20s/it, lr: 1.0e-04 loss: 5.746e-02]
anha:  89%|########9 | 2946/3300 [33:44<10:58,  1.86s/it, lr: 1.0e-04 loss: 5.746e-02]
anha:  89%|########9 | 2946/3300 [33:44<10:58,  1.86s/it, lr: 1.0e-04 loss: 5.746e-02]
anha:  89%|########9 | 2946/3300 [33:47<10:58,  1.86s/it, lr: 1.0e-04 loss: 4.721e-02]
anha:  89%|########9 | 2946/3300 [33:47<10:58,  1.86s/it, lr: 1.0e-04 loss: 4.721e-02]
anha:  89%|########9 | 2947/3300 [33:47<13:25,  2.28s/it, lr: 1.0e-04 loss: 4.721e-02]
anha:  89%|########9 | 2947/3300 [33:47<13:25,  2.28s/it, lr: 1.0e-04 loss: 4.721e-02]
anha:  89%|########9 | 2947/3300 [33:48<13:25,  2.28s/it, lr: 1.0e-04 loss: 7.059e-02]
anha:  89%|########9 | 2947/3300 [33:48<13:25,  2.28s/it, lr: 1.0e-04 loss: 7.059e-02]
anha:  89%|########9 | 2948/3300 [33:48<11:10,  1.91s/it, lr: 1.0e-04 loss: 7.059e-02]
anha:  89%|########9 | 2948/3300 [33:48<11:10,  1.91s/it, lr: 1.0e-04 loss: 7.059e-02]
anha:  89%|########9 | 2948/3300 [33:49<11:10,  1.91s/it, lr: 1.0e-04 loss: 3.826e-02]
anha:  89%|########9 | 2948/3300 [33:49<11:10,  1.91s/it, lr: 1.0e-04 loss: 3.826e-02]
anha:  89%|########9 | 2949/3300 [33:49<09:36,  1.64s/it, lr: 1.0e-04 loss: 3.826e-02]
anha:  89%|########9 | 2949/3300 [33:49<09:36,  1.64s/it, lr: 1.0e-04 loss: 3.826e-02]
anha:  89%|########9 | 2949/3300 [33:50<09:36,  1.64s/it, lr: 1.0e-04 loss: 1.957e-01]
anha:  89%|########9 | 2949/3300 [33:50<09:36,  1.64s/it, lr: 1.0e-04 loss: 1.957e-01]
anha:  89%|########9 | 2950/3300 [33:53<09:34,  1.64s/it, lr: 1.0e-04 loss: 3.510e-02]
anha:  89%|########9 | 2950/3300 [33:53<09:34,  1.64s/it, lr: 1.0e-04 loss: 3.510e-02]
anha:  89%|########9 | 2951/3300 [33:53<09:22,  1.61s/it, lr: 1.0e-04 loss: 3.510e-02]
anha:  89%|########9 | 2951/3300 [33:53<09:22,  1.61s/it, lr: 1.0e-04 loss: 3.510e-02]
anha:  89%|########9 | 2951/3300 [33:54<09:22,  1.61s/it, lr: 1.0e-04 loss: 4.703e-02]
anha:  89%|########9 | 2951/3300 [33:54<09:22,  1.61s/it, lr: 1.0e-04 loss: 4.703e-02]
anha:  89%|########9 | 2952/3300 [33:54<08:28,  1.46s/it, lr: 1.0e-04 loss: 4.703e-02]
anha:  89%|########9 | 2952/3300 [33:54<08:28,  1.46s/it, lr: 1.0e-04 loss: 4.703e-02]
anha:  89%|########9 | 2952/3300 [33:55<08:28,  1.46s/it, lr: 1.0e-04 loss: 7.060e-02]
anha:  89%|########9 | 2952/3300 [33:55<08:28,  1.46s/it, lr: 1.0e-04 loss: 7.060e-02]
anha:  89%|########9 | 2953/3300 [33:55<07:45,  1.34s/it, lr: 1.0e-04 loss: 7.060e-02]
anha:  89%|########9 | 2953/3300 [33:55<07:45,  1.34s/it, lr: 1.0e-04 loss: 7.060e-02]
anha:  89%|########9 | 2953/3300 [33:56<07:45,  1.34s/it, lr: 1.0e-04 loss: 8.310e-02]
anha:  89%|########9 | 2953/3300 [33:56<07:45,  1.34s/it, lr: 1.0e-04 loss: 8.310e-02]
anha:  90%|########9 | 2954/3300 [33:56<07:12,  1.25s/it, lr: 1.0e-04 loss: 8.310e-02]
anha:  90%|########9 | 2954/3300 [33:56<07:12,  1.25s/it, lr: 1.0e-04 loss: 8.310e-02]
anha:  90%|########9 | 2954/3300 [33:59<07:12,  1.25s/it, lr: 1.0e-04 loss: 1.279e-01]
anha:  90%|########9 | 2954/3300 [33:59<07:12,  1.25s/it, lr: 1.0e-04 loss: 1.279e-01]
anha:  90%|########9 | 2955/3300 [33:59<10:46,  1.87s/it, lr: 1.0e-04 loss: 1.279e-01]
anha:  90%|########9 | 2955/3300 [33:59<10:46,  1.87s/it, lr: 1.0e-04 loss: 1.279e-01]
anha:  90%|########9 | 2955/3300 [34:01<10:46,  1.87s/it, lr: 1.0e-04 loss: 5.920e-02]
anha:  90%|########9 | 2955/3300 [34:01<10:46,  1.87s/it, lr: 1.0e-04 loss: 5.920e-02]
anha:  90%|########9 | 2956/3300 [34:01<10:53,  1.90s/it, lr: 1.0e-04 loss: 5.920e-02]
anha:  90%|########9 | 2956/3300 [34:01<10:53,  1.90s/it, lr: 1.0e-04 loss: 5.920e-02]
anha:  90%|########9 | 2956/3300 [34:03<10:53,  1.90s/it, lr: 1.0e-04 loss: 4.972e-02]
anha:  90%|########9 | 2956/3300 [34:03<10:53,  1.90s/it, lr: 1.0e-04 loss: 4.972e-02]
anha:  90%|########9 | 2957/3300 [34:03<10:43,  1.88s/it, lr: 1.0e-04 loss: 4.972e-02]
anha:  90%|########9 | 2957/3300 [34:03<10:43,  1.88s/it, lr: 1.0e-04 loss: 4.972e-02]
anha:  90%|########9 | 2957/3300 [34:04<10:43,  1.88s/it, lr: 1.0e-04 loss: 7.616e-02]
anha:  90%|########9 | 2957/3300 [34:04<10:43,  1.88s/it, lr: 1.0e-04 loss: 7.616e-02]
anha:  90%|########9 | 2958/3300 [34:04<09:14,  1.62s/it, lr: 1.0e-04 loss: 7.616e-02]
anha:  90%|########9 | 2958/3300 [34:04<09:14,  1.62s/it, lr: 1.0e-04 loss: 7.616e-02]
anha:  90%|########9 | 2958/3300 [34:06<09:14,  1.62s/it, lr: 1.0e-04 loss: 3.722e-02]
anha:  90%|########9 | 2958/3300 [34:06<09:14,  1.62s/it, lr: 1.0e-04 loss: 3.722e-02]
anha:  90%|########9 | 2959/3300 [34:06<09:46,  1.72s/it, lr: 1.0e-04 loss: 3.722e-02]
anha:  90%|########9 | 2959/3300 [34:06<09:46,  1.72s/it, lr: 1.0e-04 loss: 3.722e-02]
anha:  90%|########9 | 2959/3300 [34:09<09:46,  1.72s/it, lr: 1.0e-04 loss: 4.237e-02]
anha:  90%|########9 | 2959/3300 [34:09<09:46,  1.72s/it, lr: 1.0e-04 loss: 4.237e-02]
anha:  90%|########9 | 2960/3300 [34:11<09:44,  1.72s/it, lr: 1.0e-04 loss: 4.674e-02]
anha:  90%|########9 | 2960/3300 [34:11<09:44,  1.72s/it, lr: 1.0e-04 loss: 4.674e-02]
anha:  90%|########9 | 2961/3300 [34:11<07:47,  1.38s/it, lr: 1.0e-04 loss: 4.674e-02]
anha:  90%|########9 | 2961/3300 [34:11<07:47,  1.38s/it, lr: 1.0e-04 loss: 4.674e-02]
anha:  90%|########9 | 2961/3300 [34:13<07:47,  1.38s/it, lr: 1.0e-04 loss: 5.113e-02]
anha:  90%|########9 | 2961/3300 [34:13<07:47,  1.38s/it, lr: 1.0e-04 loss: 5.113e-02]
anha:  90%|########9 | 2962/3300 [34:13<08:36,  1.53s/it, lr: 1.0e-04 loss: 5.113e-02]
anha:  90%|########9 | 2962/3300 [34:13<08:36,  1.53s/it, lr: 1.0e-04 loss: 5.113e-02]
anha:  90%|########9 | 2962/3300 [34:16<08:36,  1.53s/it, lr: 1.0e-04 loss: 3.123e-02]
anha:  90%|########9 | 2962/3300 [34:16<08:36,  1.53s/it, lr: 1.0e-04 loss: 3.123e-02]
anha:  90%|########9 | 2963/3300 [34:16<11:16,  2.01s/it, lr: 1.0e-04 loss: 3.123e-02]
anha:  90%|########9 | 2963/3300 [34:16<11:16,  2.01s/it, lr: 1.0e-04 loss: 3.123e-02]
anha:  90%|########9 | 2963/3300 [34:18<11:16,  2.01s/it, lr: 1.0e-04 loss: 5.903e-02]
anha:  90%|########9 | 2963/3300 [34:18<11:16,  2.01s/it, lr: 1.0e-04 loss: 5.903e-02]
anha:  90%|########9 | 2964/3300 [34:18<10:48,  1.93s/it, lr: 1.0e-04 loss: 5.903e-02]
anha:  90%|########9 | 2964/3300 [34:18<10:48,  1.93s/it, lr: 1.0e-04 loss: 5.903e-02]
anha:  90%|########9 | 2964/3300 [34:20<10:48,  1.93s/it, lr: 1.0e-04 loss: 4.436e-02]
anha:  90%|########9 | 2964/3300 [34:20<10:48,  1.93s/it, lr: 1.0e-04 loss: 4.436e-02]
anha:  90%|########9 | 2965/3300 [34:20<10:50,  1.94s/it, lr: 1.0e-04 loss: 4.436e-02]
anha:  90%|########9 | 2965/3300 [34:20<10:50,  1.94s/it, lr: 1.0e-04 loss: 4.436e-02]
anha:  90%|########9 | 2965/3300 [34:21<10:50,  1.94s/it, lr: 1.0e-04 loss: 1.231e-01]
anha:  90%|########9 | 2965/3300 [34:21<10:50,  1.94s/it, lr: 1.0e-04 loss: 1.231e-01]
anha:  90%|########9 | 2966/3300 [34:21<09:28,  1.70s/it, lr: 1.0e-04 loss: 1.231e-01]
anha:  90%|########9 | 2966/3300 [34:21<09:28,  1.70s/it, lr: 1.0e-04 loss: 1.231e-01]
anha:  90%|########9 | 2966/3300 [34:23<09:28,  1.70s/it, lr: 1.0e-04 loss: 3.215e-02]
anha:  90%|########9 | 2966/3300 [34:23<09:28,  1.70s/it, lr: 1.0e-04 loss: 3.215e-02]
anha:  90%|########9 | 2967/3300 [34:23<09:27,  1.71s/it, lr: 1.0e-04 loss: 3.215e-02]
anha:  90%|########9 | 2967/3300 [34:23<09:27,  1.71s/it, lr: 1.0e-04 loss: 3.215e-02]
anha:  90%|########9 | 2967/3300 [34:24<09:27,  1.71s/it, lr: 1.0e-04 loss: 9.314e-02]
anha:  90%|########9 | 2967/3300 [34:24<09:27,  1.71s/it, lr: 1.0e-04 loss: 9.314e-02]
anha:  90%|########9 | 2968/3300 [34:24<08:16,  1.50s/it, lr: 1.0e-04 loss: 9.314e-02]
anha:  90%|########9 | 2968/3300 [34:24<08:16,  1.50s/it, lr: 1.0e-04 loss: 9.314e-02]
anha:  90%|########9 | 2968/3300 [34:27<08:16,  1.50s/it, lr: 1.0e-04 loss: 1.251e-01]
anha:  90%|########9 | 2968/3300 [34:27<08:16,  1.50s/it, lr: 1.0e-04 loss: 1.251e-01]
anha:  90%|########9 | 2969/3300 [34:27<11:28,  2.08s/it, lr: 1.0e-04 loss: 1.251e-01]
anha:  90%|########9 | 2969/3300 [34:27<11:28,  2.08s/it, lr: 1.0e-04 loss: 1.251e-01]
anha:  90%|########9 | 2969/3300 [34:29<11:28,  2.08s/it, lr: 1.0e-04 loss: 5.151e-02]
anha:  90%|########9 | 2969/3300 [34:29<11:28,  2.08s/it, lr: 1.0e-04 loss: 5.151e-02]
Saving at step 2970
Saved checkpoint to /workspace/ai-toolkit/output/anha/anha_000002970.safetensors
Saved optimizer to /workspace/ai-toolkit/output/anha/optimizer.pt

anha:  90%|######### | 2970/3300 [34:32<11:26,  2.08s/it, lr: 1.0e-04 loss: 1.743e-01]
anha:  90%|######### | 2970/3300 [34:32<11:26,  2.08s/it, lr: 1.0e-04 loss: 1.743e-01]
anha:  90%|######### | 2971/3300 [34:32<10:28,  1.91s/it, lr: 1.0e-04 loss: 1.743e-01]
anha:  90%|######### | 2971/3300 [34:32<10:28,  1.91s/it, lr: 1.0e-04 loss: 1.743e-01]
anha:  90%|######### | 2971/3300 [34:36<10:28,  1.91s/it, lr: 1.0e-04 loss: 5.102e-02]
anha:  90%|######### | 2971/3300 [34:36<10:28,  1.91s/it, lr: 1.0e-04 loss: 5.102e-02]
anha:  90%|######### | 2972/3300 [34:36<12:34,  2.30s/it, lr: 1.0e-04 loss: 5.102e-02]
anha:  90%|######### | 2972/3300 [34:36<12:34,  2.30s/it, lr: 1.0e-04 loss: 5.102e-02]
anha:  90%|######### | 2972/3300 [34:39<12:34,  2.30s/it, lr: 1.0e-04 loss: 3.335e-02]
anha:  90%|######### | 2972/3300 [34:39<12:34,  2.30s/it, lr: 1.0e-04 loss: 3.335e-02]
anha:  90%|######### | 2973/3300 [34:39<14:04,  2.58s/it, lr: 1.0e-04 loss: 3.335e-02]
anha:  90%|######### | 2973/3300 [34:39<14:04,  2.58s/it, lr: 1.0e-04 loss: 3.335e-02]
anha:  90%|######### | 2973/3300 [34:40<14:04,  2.58s/it, lr: 1.0e-04 loss: 3.605e-02]
anha:  90%|######### | 2973/3300 [34:40<14:04,  2.58s/it, lr: 1.0e-04 loss: 3.605e-02]
anha:  90%|######### | 2974/3300 [34:40<11:42,  2.16s/it, lr: 1.0e-04 loss: 3.605e-02]
anha:  90%|######### | 2974/3300 [34:40<11:42,  2.16s/it, lr: 1.0e-04 loss: 3.605e-02]
anha:  90%|######### | 2974/3300 [34:42<11:42,  2.16s/it, lr: 1.0e-04 loss: 1.027e-01]
anha:  90%|######### | 2974/3300 [34:42<11:42,  2.16s/it, lr: 1.0e-04 loss: 1.027e-01]
anha:  90%|######### | 2975/3300 [34:42<11:22,  2.10s/it, lr: 1.0e-04 loss: 1.027e-01]
anha:  90%|######### | 2975/3300 [34:42<11:22,  2.10s/it, lr: 1.0e-04 loss: 1.027e-01]
anha:  90%|######### | 2975/3300 [34:43<11:22,  2.10s/it, lr: 1.0e-04 loss: 5.419e-02]
anha:  90%|######### | 2975/3300 [34:43<11:22,  2.10s/it, lr: 1.0e-04 loss: 5.419e-02]
anha:  90%|######### | 2976/3300 [34:43<09:39,  1.79s/it, lr: 1.0e-04 loss: 5.419e-02]
anha:  90%|######### | 2976/3300 [34:43<09:39,  1.79s/it, lr: 1.0e-04 loss: 5.419e-02]
anha:  90%|######### | 2976/3300 [34:47<09:39,  1.79s/it, lr: 1.0e-04 loss: 1.820e-01]
anha:  90%|######### | 2976/3300 [34:47<09:39,  1.79s/it, lr: 1.0e-04 loss: 1.820e-01]
anha:  90%|######### | 2977/3300 [34:47<11:57,  2.22s/it, lr: 1.0e-04 loss: 1.820e-01]
anha:  90%|######### | 2977/3300 [34:47<11:57,  2.22s/it, lr: 1.0e-04 loss: 1.820e-01]
anha:  90%|######### | 2977/3300 [34:48<11:57,  2.22s/it, lr: 1.0e-04 loss: 9.130e-02]
anha:  90%|######### | 2977/3300 [34:48<11:57,  2.22s/it, lr: 1.0e-04 loss: 9.130e-02]
anha:  90%|######### | 2978/3300 [34:48<09:59,  1.86s/it, lr: 1.0e-04 loss: 9.130e-02]
anha:  90%|######### | 2978/3300 [34:48<09:59,  1.86s/it, lr: 1.0e-04 loss: 9.130e-02]
anha:  90%|######### | 2978/3300 [34:49<09:59,  1.86s/it, lr: 1.0e-04 loss: 5.769e-02]
anha:  90%|######### | 2978/3300 [34:49<09:59,  1.86s/it, lr: 1.0e-04 loss: 5.769e-02]
anha:  90%|######### | 2979/3300 [34:49<08:46,  1.64s/it, lr: 1.0e-04 loss: 5.769e-02]
anha:  90%|######### | 2979/3300 [34:49<08:46,  1.64s/it, lr: 1.0e-04 loss: 5.769e-02]
anha:  90%|######### | 2979/3300 [34:50<08:46,  1.64s/it, lr: 1.0e-04 loss: 9.859e-02]
anha:  90%|######### | 2979/3300 [34:50<08:46,  1.64s/it, lr: 1.0e-04 loss: 9.859e-02]
anha:  90%|######### | 2980/3300 [34:53<08:44,  1.64s/it, lr: 1.0e-04 loss: 3.849e-02]
anha:  90%|######### | 2980/3300 [34:53<08:44,  1.64s/it, lr: 1.0e-04 loss: 3.849e-02]
anha:  90%|######### | 2981/3300 [34:53<08:54,  1.68s/it, lr: 1.0e-04 loss: 3.849e-02]
anha:  90%|######### | 2981/3300 [34:53<08:54,  1.68s/it, lr: 1.0e-04 loss: 3.849e-02]
anha:  90%|######### | 2981/3300 [34:54<08:54,  1.68s/it, lr: 1.0e-04 loss: 7.765e-02]
anha:  90%|######### | 2981/3300 [34:54<08:54,  1.68s/it, lr: 1.0e-04 loss: 7.765e-02]
anha:  90%|######### | 2982/3300 [34:54<08:00,  1.51s/it, lr: 1.0e-04 loss: 7.765e-02]
anha:  90%|######### | 2982/3300 [34:54<08:00,  1.51s/it, lr: 1.0e-04 loss: 7.765e-02]
anha:  90%|######### | 2982/3300 [34:55<08:00,  1.51s/it, lr: 1.0e-04 loss: 9.331e-02]
anha:  90%|######### | 2982/3300 [34:55<08:00,  1.51s/it, lr: 1.0e-04 loss: 9.331e-02]
anha:  90%|######### | 2983/3300 [34:55<07:17,  1.38s/it, lr: 1.0e-04 loss: 9.331e-02]
anha:  90%|######### | 2983/3300 [34:55<07:17,  1.38s/it, lr: 1.0e-04 loss: 9.331e-02]
anha:  90%|######### | 2983/3300 [34:59<07:17,  1.38s/it, lr: 1.0e-04 loss: 2.907e-02]
anha:  90%|######### | 2983/3300 [34:59<07:17,  1.38s/it, lr: 1.0e-04 loss: 2.907e-02]
anha:  90%|######### | 2984/3300 [34:59<10:06,  1.92s/it, lr: 1.0e-04 loss: 2.907e-02]
anha:  90%|######### | 2984/3300 [34:59<10:06,  1.92s/it, lr: 1.0e-04 loss: 2.907e-02]
anha:  90%|######### | 2984/3300 [34:59<10:06,  1.92s/it, lr: 1.0e-04 loss: 3.586e-02]
anha:  90%|######### | 2984/3300 [34:59<10:06,  1.92s/it, lr: 1.0e-04 loss: 3.586e-02]
anha:  90%|######### | 2985/3300 [34:59<08:43,  1.66s/it, lr: 1.0e-04 loss: 3.586e-02]
anha:  90%|######### | 2985/3300 [34:59<08:43,  1.66s/it, lr: 1.0e-04 loss: 3.586e-02]
anha:  90%|######### | 2985/3300 [35:03<08:43,  1.66s/it, lr: 1.0e-04 loss: 4.034e-02]
anha:  90%|######### | 2985/3300 [35:03<08:43,  1.66s/it, lr: 1.0e-04 loss: 4.034e-02]
anha:  90%|######### | 2986/3300 [35:03<11:14,  2.15s/it, lr: 1.0e-04 loss: 4.034e-02]
anha:  90%|######### | 2986/3300 [35:03<11:14,  2.15s/it, lr: 1.0e-04 loss: 4.034e-02]
anha:  90%|######### | 2986/3300 [35:05<11:14,  2.15s/it, lr: 1.0e-04 loss: 4.979e-02]
anha:  90%|######### | 2986/3300 [35:05<11:14,  2.15s/it, lr: 1.0e-04 loss: 4.979e-02]
anha:  91%|######### | 2987/3300 [35:05<10:35,  2.03s/it, lr: 1.0e-04 loss: 4.979e-02]
anha:  91%|######### | 2987/3300 [35:05<10:35,  2.03s/it, lr: 1.0e-04 loss: 4.979e-02]
anha:  91%|######### | 2987/3300 [35:06<10:35,  2.03s/it, lr: 1.0e-04 loss: 3.043e-02]
anha:  91%|######### | 2987/3300 [35:06<10:35,  2.03s/it, lr: 1.0e-04 loss: 3.043e-02]
anha:  91%|######### | 2988/3300 [35:06<09:09,  1.76s/it, lr: 1.0e-04 loss: 3.043e-02]
anha:  91%|######### | 2988/3300 [35:06<09:09,  1.76s/it, lr: 1.0e-04 loss: 3.043e-02]
anha:  91%|######### | 2988/3300 [35:08<09:09,  1.76s/it, lr: 1.0e-04 loss: 1.551e-01]
anha:  91%|######### | 2988/3300 [35:08<09:09,  1.76s/it, lr: 1.0e-04 loss: 1.551e-01]
anha:  91%|######### | 2989/3300 [35:08<09:27,  1.83s/it, lr: 1.0e-04 loss: 1.551e-01]
anha:  91%|######### | 2989/3300 [35:08<09:27,  1.83s/it, lr: 1.0e-04 loss: 1.551e-01]
anha:  91%|######### | 2989/3300 [35:11<09:27,  1.83s/it, lr: 1.0e-04 loss: 3.486e-01]
anha:  91%|######### | 2989/3300 [35:11<09:27,  1.83s/it, lr: 1.0e-04 loss: 3.486e-01]
anha:  91%|######### | 2990/3300 [35:13<09:26,  1.83s/it, lr: 1.0e-04 loss: 6.414e-02]
anha:  91%|######### | 2990/3300 [35:13<09:26,  1.83s/it, lr: 1.0e-04 loss: 6.414e-02]
anha:  91%|######### | 2991/3300 [35:13<07:08,  1.39s/it, lr: 1.0e-04 loss: 6.414e-02]
anha:  91%|######### | 2991/3300 [35:13<07:08,  1.39s/it, lr: 1.0e-04 loss: 6.414e-02]
anha:  91%|######### | 2991/3300 [35:14<07:08,  1.39s/it, lr: 1.0e-04 loss: 4.554e-02]
anha:  91%|######### | 2991/3300 [35:14<07:08,  1.39s/it, lr: 1.0e-04 loss: 4.554e-02]
anha:  91%|######### | 2992/3300 [35:14<06:38,  1.29s/it, lr: 1.0e-04 loss: 4.554e-02]
anha:  91%|######### | 2992/3300 [35:14<06:38,  1.29s/it, lr: 1.0e-04 loss: 4.554e-02]
anha:  91%|######### | 2992/3300 [35:17<06:38,  1.29s/it, lr: 1.0e-04 loss: 1.757e-01]
anha:  91%|######### | 2992/3300 [35:17<06:38,  1.29s/it, lr: 1.0e-04 loss: 1.757e-01]
anha:  91%|######### | 2993/3300 [35:17<09:12,  1.80s/it, lr: 1.0e-04 loss: 1.757e-01]
anha:  91%|######### | 2993/3300 [35:17<09:12,  1.80s/it, lr: 1.0e-04 loss: 1.757e-01]
anha:  91%|######### | 2993/3300 [35:19<09:12,  1.80s/it, lr: 1.0e-04 loss: 1.016e-01]
anha:  91%|######### | 2993/3300 [35:19<09:12,  1.80s/it, lr: 1.0e-04 loss: 1.016e-01]
anha:  91%|######### | 2994/3300 [35:19<09:25,  1.85s/it, lr: 1.0e-04 loss: 1.016e-01]
anha:  91%|######### | 2994/3300 [35:19<09:25,  1.85s/it, lr: 1.0e-04 loss: 1.016e-01]
anha:  91%|######### | 2994/3300 [35:20<09:25,  1.85s/it, lr: 1.0e-04 loss: 8.708e-02]
anha:  91%|######### | 2994/3300 [35:20<09:25,  1.85s/it, lr: 1.0e-04 loss: 8.708e-02]
anha:  91%|######### | 2995/3300 [35:20<08:13,  1.62s/it, lr: 1.0e-04 loss: 8.708e-02]
anha:  91%|######### | 2995/3300 [35:20<08:13,  1.62s/it, lr: 1.0e-04 loss: 8.708e-02]
anha:  91%|######### | 2995/3300 [35:21<08:13,  1.62s/it, lr: 1.0e-04 loss: 5.290e-02]
anha:  91%|######### | 2995/3300 [35:21<08:13,  1.62s/it, lr: 1.0e-04 loss: 5.290e-02]
anha:  91%|######### | 2996/3300 [35:21<07:18,  1.44s/it, lr: 1.0e-04 loss: 5.290e-02]
anha:  91%|######### | 2996/3300 [35:21<07:18,  1.44s/it, lr: 1.0e-04 loss: 5.290e-02]
anha:  91%|######### | 2996/3300 [35:25<07:18,  1.44s/it, lr: 1.0e-04 loss: 2.832e-02]
anha:  91%|######### | 2996/3300 [35:25<07:18,  1.44s/it, lr: 1.0e-04 loss: 2.832e-02]
anha:  91%|######### | 2997/3300 [35:25<10:08,  2.01s/it, lr: 1.0e-04 loss: 2.832e-02]
anha:  91%|######### | 2997/3300 [35:25<10:08,  2.01s/it, lr: 1.0e-04 loss: 2.832e-02]
anha:  91%|######### | 2997/3300 [35:26<10:08,  2.01s/it, lr: 1.0e-04 loss: 4.123e-02]
anha:  91%|######### | 2997/3300 [35:26<10:08,  2.01s/it, lr: 1.0e-04 loss: 4.123e-02]
anha:  91%|######### | 2998/3300 [35:26<08:38,  1.72s/it, lr: 1.0e-04 loss: 4.123e-02]
anha:  91%|######### | 2998/3300 [35:26<08:38,  1.72s/it, lr: 1.0e-04 loss: 4.123e-02]
anha:  91%|######### | 2998/3300 [35:27<08:38,  1.72s/it, lr: 1.0e-04 loss: 3.718e-02]
anha:  91%|######### | 2998/3300 [35:27<08:38,  1.72s/it, lr: 1.0e-04 loss: 3.718e-02]
anha:  91%|######### | 2999/3300 [35:27<08:38,  1.72s/it, lr: 1.0e-04 loss: 3.718e-02]
anha:  91%|######### | 2999/3300 [35:27<08:38,  1.72s/it, lr: 1.0e-04 loss: 3.718e-02]
anha:  91%|######### | 2999/3300 [35:29<08:38,  1.72s/it, lr: 1.0e-04 loss: 1.616e-01]
anha:  91%|######### | 2999/3300 [35:29<08:38,  1.72s/it, lr: 1.0e-04 loss: 1.616e-01]


Generating Images: 0it [00:00, ?it/s]
Generating Images: 0it [00:00, ?it/s]


                                     
                                     
anha:  91%|######### | 3000/3300 [35:31<08:37,  1.72s/it, lr: 1.0e-04 loss: 5.828e-02]
anha:  91%|######### | 3000/3300 [35:31<08:37,  1.72s/it, lr: 1.0e-04 loss: 5.828e-02]
anha:  91%|######### | 3001/3300 [35:31<06:53,  1.38s/it, lr: 1.0e-04 loss: 5.828e-02]
anha:  91%|######### | 3001/3300 [35:31<06:53,  1.38s/it, lr: 1.0e-04 loss: 5.828e-02]
anha:  91%|######### | 3001/3300 [35:35<06:53,  1.38s/it, lr: 1.0e-04 loss: 4.864e-02]
anha:  91%|######### | 3001/3300 [35:35<06:53,  1.38s/it, lr: 1.0e-04 loss: 4.864e-02]
anha:  91%|######### | 3002/3300 [35:35<09:25,  1.90s/it, lr: 1.0e-04 loss: 4.864e-02]
anha:  91%|######### | 3002/3300 [35:35<09:25,  1.90s/it, lr: 1.0e-04 loss: 4.864e-02]
anha:  91%|######### | 3002/3300 [35:38<09:25,  1.90s/it, lr: 1.0e-04 loss: 9.256e-02]
anha:  91%|######### | 3002/3300 [35:38<09:25,  1.90s/it, lr: 1.0e-04 loss: 9.256e-02]
anha:  91%|#########1| 3003/3300 [35:38<11:25,  2.31s/it, lr: 1.0e-04 loss: 9.256e-02]
anha:  91%|#########1| 3003/3300 [35:38<11:25,  2.31s/it, lr: 1.0e-04 loss: 9.256e-02]
anha:  91%|#########1| 3003/3300 [35:40<11:25,  2.31s/it, lr: 1.0e-04 loss: 4.286e-02]
anha:  91%|#########1| 3003/3300 [35:40<11:25,  2.31s/it, lr: 1.0e-04 loss: 4.286e-02]
anha:  91%|#########1| 3004/3300 [35:40<10:35,  2.15s/it, lr: 1.0e-04 loss: 4.286e-02]
anha:  91%|#########1| 3004/3300 [35:40<10:35,  2.15s/it, lr: 1.0e-04 loss: 4.286e-02]
anha:  91%|#########1| 3004/3300 [35:41<10:35,  2.15s/it, lr: 1.0e-04 loss: 7.696e-02]
anha:  91%|#########1| 3004/3300 [35:41<10:35,  2.15s/it, lr: 1.0e-04 loss: 7.696e-02]
anha:  91%|#########1| 3005/3300 [35:41<08:59,  1.83s/it, lr: 1.0e-04 loss: 7.696e-02]
anha:  91%|#########1| 3005/3300 [35:41<08:59,  1.83s/it, lr: 1.0e-04 loss: 7.696e-02]
anha:  91%|#########1| 3005/3300 [35:44<08:59,  1.83s/it, lr: 1.0e-04 loss: 9.882e-02]
anha:  91%|#########1| 3005/3300 [35:44<08:59,  1.83s/it, lr: 1.0e-04 loss: 9.882e-02]
anha:  91%|#########1| 3006/3300 [35:44<11:00,  2.25s/it, lr: 1.0e-04 loss: 9.882e-02]
anha:  91%|#########1| 3006/3300 [35:44<11:00,  2.25s/it, lr: 1.0e-04 loss: 9.882e-02]
anha:  91%|#########1| 3006/3300 [35:45<11:00,  2.25s/it, lr: 1.0e-04 loss: 2.826e-01]
anha:  91%|#########1| 3006/3300 [35:45<11:00,  2.25s/it, lr: 1.0e-04 loss: 2.826e-01]
anha:  91%|#########1| 3007/3300 [35:45<09:13,  1.89s/it, lr: 1.0e-04 loss: 2.826e-01]
anha:  91%|#########1| 3007/3300 [35:45<09:13,  1.89s/it, lr: 1.0e-04 loss: 2.826e-01]
anha:  91%|#########1| 3007/3300 [35:46<09:13,  1.89s/it, lr: 1.0e-04 loss: 6.714e-02]
anha:  91%|#########1| 3007/3300 [35:46<09:13,  1.89s/it, lr: 1.0e-04 loss: 6.714e-02]
anha:  91%|#########1| 3008/3300 [35:46<07:54,  1.63s/it, lr: 1.0e-04 loss: 6.714e-02]
anha:  91%|#########1| 3008/3300 [35:46<07:54,  1.63s/it, lr: 1.0e-04 loss: 6.714e-02]
anha:  91%|#########1| 3008/3300 [35:48<07:54,  1.63s/it, lr: 1.0e-04 loss: 4.796e-02]
anha:  91%|#########1| 3008/3300 [35:48<07:54,  1.63s/it, lr: 1.0e-04 loss: 4.796e-02]
anha:  91%|#########1| 3009/3300 [35:48<08:21,  1.72s/it, lr: 1.0e-04 loss: 4.796e-02]
anha:  91%|#########1| 3009/3300 [35:48<08:21,  1.72s/it, lr: 1.0e-04 loss: 4.796e-02]
anha:  91%|#########1| 3009/3300 [35:52<08:21,  1.72s/it, lr: 1.0e-04 loss: 7.958e-02]
anha:  91%|#########1| 3009/3300 [35:52<08:21,  1.72s/it, lr: 1.0e-04 loss: 7.958e-02]
anha:  91%|#########1| 3010/3300 [35:54<08:19,  1.72s/it, lr: 1.0e-04 loss: 2.571e-02]
anha:  91%|#########1| 3010/3300 [35:54<08:19,  1.72s/it, lr: 1.0e-04 loss: 2.571e-02]
anha:  91%|#########1| 3011/3300 [35:54<06:39,  1.38s/it, lr: 1.0e-04 loss: 2.571e-02]
anha:  91%|#########1| 3011/3300 [35:54<06:39,  1.38s/it, lr: 1.0e-04 loss: 2.571e-02]
anha:  91%|#########1| 3011/3300 [35:57<06:39,  1.38s/it, lr: 1.0e-04 loss: 4.773e-02]
anha:  91%|#########1| 3011/3300 [35:57<06:39,  1.38s/it, lr: 1.0e-04 loss: 4.773e-02]
anha:  91%|#########1| 3012/3300 [35:57<09:00,  1.88s/it, lr: 1.0e-04 loss: 4.773e-02]
anha:  91%|#########1| 3012/3300 [35:57<09:00,  1.88s/it, lr: 1.0e-04 loss: 4.773e-02]
anha:  91%|#########1| 3012/3300 [35:59<09:00,  1.88s/it, lr: 1.0e-04 loss: 8.361e-02]
anha:  91%|#########1| 3012/3300 [35:59<09:00,  1.88s/it, lr: 1.0e-04 loss: 8.361e-02]
anha:  91%|#########1| 3013/3300 [35:59<09:06,  1.90s/it, lr: 1.0e-04 loss: 8.361e-02]
anha:  91%|#########1| 3013/3300 [35:59<09:06,  1.90s/it, lr: 1.0e-04 loss: 8.361e-02]
anha:  91%|#########1| 3013/3300 [36:02<09:06,  1.90s/it, lr: 1.0e-04 loss: 9.608e-02]
anha:  91%|#########1| 3013/3300 [36:02<09:06,  1.90s/it, lr: 1.0e-04 loss: 9.608e-02]
anha:  91%|#########1| 3014/3300 [36:02<10:53,  2.28s/it, lr: 1.0e-04 loss: 9.608e-02]
anha:  91%|#########1| 3014/3300 [36:02<10:53,  2.28s/it, lr: 1.0e-04 loss: 9.608e-02]
anha:  91%|#########1| 3014/3300 [36:03<10:53,  2.28s/it, lr: 1.0e-04 loss: 9.829e-02]
anha:  91%|#########1| 3014/3300 [36:03<10:53,  2.28s/it, lr: 1.0e-04 loss: 9.829e-02]
anha:  91%|#########1| 3015/3300 [36:03<09:09,  1.93s/it, lr: 1.0e-04 loss: 9.829e-02]
anha:  91%|#########1| 3015/3300 [36:03<09:09,  1.93s/it, lr: 1.0e-04 loss: 9.829e-02]
anha:  91%|#########1| 3015/3300 [36:05<09:09,  1.93s/it, lr: 1.0e-04 loss: 3.108e-02]
anha:  91%|#########1| 3015/3300 [36:05<09:09,  1.93s/it, lr: 1.0e-04 loss: 3.108e-02]
anha:  91%|#########1| 3016/3300 [36:05<09:11,  1.94s/it, lr: 1.0e-04 loss: 3.108e-02]
anha:  91%|#########1| 3016/3300 [36:05<09:11,  1.94s/it, lr: 1.0e-04 loss: 3.108e-02]
anha:  91%|#########1| 3016/3300 [36:07<09:11,  1.94s/it, lr: 1.0e-04 loss: 4.203e-02]
anha:  91%|#########1| 3016/3300 [36:07<09:11,  1.94s/it, lr: 1.0e-04 loss: 4.203e-02]
anha:  91%|#########1| 3017/3300 [36:07<09:11,  1.95s/it, lr: 1.0e-04 loss: 4.203e-02]
anha:  91%|#########1| 3017/3300 [36:07<09:11,  1.95s/it, lr: 1.0e-04 loss: 4.203e-02]
anha:  91%|#########1| 3017/3300 [36:09<09:11,  1.95s/it, lr: 1.0e-04 loss: 5.964e-02]
anha:  91%|#########1| 3017/3300 [36:09<09:11,  1.95s/it, lr: 1.0e-04 loss: 5.964e-02]
anha:  91%|#########1| 3018/3300 [36:09<09:12,  1.96s/it, lr: 1.0e-04 loss: 5.964e-02]
anha:  91%|#########1| 3018/3300 [36:09<09:12,  1.96s/it, lr: 1.0e-04 loss: 5.964e-02]
anha:  91%|#########1| 3018/3300 [36:10<09:12,  1.96s/it, lr: 1.0e-04 loss: 3.973e-02]
anha:  91%|#########1| 3018/3300 [36:10<09:12,  1.96s/it, lr: 1.0e-04 loss: 3.973e-02]
anha:  91%|#########1| 3019/3300 [36:10<08:01,  1.71s/it, lr: 1.0e-04 loss: 3.973e-02]
anha:  91%|#########1| 3019/3300 [36:10<08:01,  1.71s/it, lr: 1.0e-04 loss: 3.973e-02]
anha:  91%|#########1| 3019/3300 [36:12<08:01,  1.71s/it, lr: 1.0e-04 loss: 6.793e-02]
anha:  91%|#########1| 3019/3300 [36:12<08:01,  1.71s/it, lr: 1.0e-04 loss: 6.793e-02]
anha:  92%|#########1| 3020/3300 [36:16<07:59,  1.71s/it, lr: 1.0e-04 loss: 1.932e-01]
anha:  92%|#########1| 3020/3300 [36:16<07:59,  1.71s/it, lr: 1.0e-04 loss: 1.932e-01]
anha:  92%|#########1| 3021/3300 [36:16<07:54,  1.70s/it, lr: 1.0e-04 loss: 1.932e-01]
anha:  92%|#########1| 3021/3300 [36:16<07:54,  1.70s/it, lr: 1.0e-04 loss: 1.932e-01]
anha:  92%|#########1| 3021/3300 [36:17<07:54,  1.70s/it, lr: 1.0e-04 loss: 8.303e-02]
anha:  92%|#########1| 3021/3300 [36:17<07:54,  1.70s/it, lr: 1.0e-04 loss: 8.303e-02]
anha:  92%|#########1| 3022/3300 [36:17<07:53,  1.70s/it, lr: 1.0e-04 loss: 8.303e-02]
anha:  92%|#########1| 3022/3300 [36:17<07:53,  1.70s/it, lr: 1.0e-04 loss: 8.303e-02]
anha:  92%|#########1| 3022/3300 [36:18<07:53,  1.70s/it, lr: 1.0e-04 loss: 7.165e-02]
anha:  92%|#########1| 3022/3300 [36:18<07:53,  1.70s/it, lr: 1.0e-04 loss: 7.165e-02]
anha:  92%|#########1| 3023/3300 [36:18<07:01,  1.52s/it, lr: 1.0e-04 loss: 7.165e-02]
anha:  92%|#########1| 3023/3300 [36:18<07:01,  1.52s/it, lr: 1.0e-04 loss: 7.165e-02]
anha:  92%|#########1| 3023/3300 [36:20<07:01,  1.52s/it, lr: 1.0e-04 loss: 8.223e-02]
anha:  92%|#########1| 3023/3300 [36:20<07:01,  1.52s/it, lr: 1.0e-04 loss: 8.223e-02]
anha:  92%|#########1| 3024/3300 [36:20<07:32,  1.64s/it, lr: 1.0e-04 loss: 8.223e-02]
anha:  92%|#########1| 3024/3300 [36:20<07:32,  1.64s/it, lr: 1.0e-04 loss: 8.223e-02]
anha:  92%|#########1| 3024/3300 [36:24<07:32,  1.64s/it, lr: 1.0e-04 loss: 1.422e-01]
anha:  92%|#########1| 3024/3300 [36:24<07:32,  1.64s/it, lr: 1.0e-04 loss: 1.422e-01]
anha:  92%|#########1| 3025/3300 [36:24<09:27,  2.06s/it, lr: 1.0e-04 loss: 1.422e-01]
anha:  92%|#########1| 3025/3300 [36:24<09:27,  2.06s/it, lr: 1.0e-04 loss: 1.422e-01]
anha:  92%|#########1| 3025/3300 [36:25<09:27,  2.06s/it, lr: 1.0e-04 loss: 4.671e-02]
anha:  92%|#########1| 3025/3300 [36:25<09:27,  2.06s/it, lr: 1.0e-04 loss: 4.671e-02]
anha:  92%|#########1| 3026/3300 [36:25<08:58,  1.96s/it, lr: 1.0e-04 loss: 4.671e-02]
anha:  92%|#########1| 3026/3300 [36:25<08:58,  1.96s/it, lr: 1.0e-04 loss: 4.671e-02]
anha:  92%|#########1| 3026/3300 [36:28<08:58,  1.96s/it, lr: 1.0e-04 loss: 3.800e-02]
anha:  92%|#########1| 3026/3300 [36:28<08:58,  1.96s/it, lr: 1.0e-04 loss: 3.800e-02]
anha:  92%|#########1| 3027/3300 [36:28<09:39,  2.12s/it, lr: 1.0e-04 loss: 3.800e-02]
anha:  92%|#########1| 3027/3300 [36:28<09:39,  2.12s/it, lr: 1.0e-04 loss: 3.800e-02]
anha:  92%|#########1| 3027/3300 [36:31<09:39,  2.12s/it, lr: 1.0e-04 loss: 3.588e-02]
anha:  92%|#########1| 3027/3300 [36:31<09:39,  2.12s/it, lr: 1.0e-04 loss: 3.588e-02]
anha:  92%|#########1| 3028/3300 [36:31<11:26,  2.53s/it, lr: 1.0e-04 loss: 3.588e-02]
anha:  92%|#########1| 3028/3300 [36:31<11:26,  2.53s/it, lr: 1.0e-04 loss: 3.588e-02]
anha:  92%|#########1| 3028/3300 [36:35<11:26,  2.53s/it, lr: 1.0e-04 loss: 5.108e-02]
anha:  92%|#########1| 3028/3300 [36:35<11:26,  2.53s/it, lr: 1.0e-04 loss: 5.108e-02]
anha:  92%|#########1| 3029/3300 [36:35<12:42,  2.81s/it, lr: 1.0e-04 loss: 5.108e-02]
anha:  92%|#########1| 3029/3300 [36:35<12:42,  2.81s/it, lr: 1.0e-04 loss: 5.108e-02]
anha:  92%|#########1| 3029/3300 [36:37<12:42,  2.81s/it, lr: 1.0e-04 loss: 3.511e-02]
anha:  92%|#########1| 3029/3300 [36:37<12:42,  2.81s/it, lr: 1.0e-04 loss: 3.511e-02]
anha:  92%|#########1| 3030/3300 [36:40<12:39,  2.81s/it, lr: 1.0e-04 loss: 3.256e-01]
anha:  92%|#########1| 3030/3300 [36:40<12:39,  2.81s/it, lr: 1.0e-04 loss: 3.256e-01]
anha:  92%|#########1| 3031/3300 [36:40<10:18,  2.30s/it, lr: 1.0e-04 loss: 3.256e-01]
anha:  92%|#########1| 3031/3300 [36:40<10:18,  2.30s/it, lr: 1.0e-04 loss: 3.256e-01]
anha:  92%|#########1| 3031/3300 [36:42<10:18,  2.30s/it, lr: 1.0e-04 loss: 5.571e-02]
anha:  92%|#########1| 3031/3300 [36:42<10:18,  2.30s/it, lr: 1.0e-04 loss: 5.571e-02]
anha:  92%|#########1| 3032/3300 [36:42<09:38,  2.16s/it, lr: 1.0e-04 loss: 5.571e-02]
anha:  92%|#########1| 3032/3300 [36:42<09:38,  2.16s/it, lr: 1.0e-04 loss: 5.571e-02]
anha:  92%|#########1| 3032/3300 [36:43<09:38,  2.16s/it, lr: 1.0e-04 loss: 6.644e-02]
anha:  92%|#########1| 3032/3300 [36:43<09:38,  2.16s/it, lr: 1.0e-04 loss: 6.644e-02]
anha:  92%|#########1| 3033/3300 [36:43<08:16,  1.86s/it, lr: 1.0e-04 loss: 6.644e-02]
anha:  92%|#########1| 3033/3300 [36:43<08:16,  1.86s/it, lr: 1.0e-04 loss: 6.644e-02]
anha:  92%|#########1| 3033/3300 [36:44<08:16,  1.86s/it, lr: 1.0e-04 loss: 4.345e-02]
anha:  92%|#########1| 3033/3300 [36:44<08:16,  1.86s/it, lr: 1.0e-04 loss: 4.345e-02]
anha:  92%|#########1| 3034/3300 [36:44<07:12,  1.63s/it, lr: 1.0e-04 loss: 4.345e-02]
anha:  92%|#########1| 3034/3300 [36:44<07:12,  1.63s/it, lr: 1.0e-04 loss: 4.345e-02]
anha:  92%|#########1| 3034/3300 [36:45<07:12,  1.63s/it, lr: 1.0e-04 loss: 2.018e-01]
anha:  92%|#########1| 3034/3300 [36:45<07:12,  1.63s/it, lr: 1.0e-04 loss: 2.018e-01]
anha:  92%|#########1| 3035/3300 [36:45<06:24,  1.45s/it, lr: 1.0e-04 loss: 2.018e-01]
anha:  92%|#########1| 3035/3300 [36:45<06:24,  1.45s/it, lr: 1.0e-04 loss: 2.018e-01]
anha:  92%|#########1| 3035/3300 [36:48<06:24,  1.45s/it, lr: 1.0e-04 loss: 6.796e-02]
anha:  92%|#########1| 3035/3300 [36:48<06:24,  1.45s/it, lr: 1.0e-04 loss: 6.796e-02]
anha:  92%|#########2| 3036/3300 [36:48<08:55,  2.03s/it, lr: 1.0e-04 loss: 6.796e-02]
anha:  92%|#########2| 3036/3300 [36:48<08:55,  2.03s/it, lr: 1.0e-04 loss: 6.796e-02]
anha:  92%|#########2| 3036/3300 [36:52<08:55,  2.03s/it, lr: 1.0e-04 loss: 6.843e-02]
anha:  92%|#########2| 3036/3300 [36:52<08:55,  2.03s/it, lr: 1.0e-04 loss: 6.843e-02]
anha:  92%|#########2| 3037/3300 [36:52<10:43,  2.45s/it, lr: 1.0e-04 loss: 6.843e-02]
anha:  92%|#########2| 3037/3300 [36:52<10:43,  2.45s/it, lr: 1.0e-04 loss: 6.843e-02]
anha:  92%|#########2| 3037/3300 [36:55<10:43,  2.45s/it, lr: 1.0e-04 loss: 1.285e-01]
anha:  92%|#########2| 3037/3300 [36:55<10:43,  2.45s/it, lr: 1.0e-04 loss: 1.285e-01]
anha:  92%|#########2| 3038/3300 [36:55<11:50,  2.71s/it, lr: 1.0e-04 loss: 1.285e-01]
anha:  92%|#########2| 3038/3300 [36:55<11:50,  2.71s/it, lr: 1.0e-04 loss: 1.285e-01]
anha:  92%|#########2| 3038/3300 [36:56<11:50,  2.71s/it, lr: 1.0e-04 loss: 1.641e-01]
anha:  92%|#########2| 3038/3300 [36:56<11:50,  2.71s/it, lr: 1.0e-04 loss: 1.641e-01]
anha:  92%|#########2| 3039/3300 [36:56<09:36,  2.21s/it, lr: 1.0e-04 loss: 1.641e-01]
anha:  92%|#########2| 3039/3300 [36:56<09:36,  2.21s/it, lr: 1.0e-04 loss: 1.641e-01]
anha:  92%|#########2| 3039/3300 [36:58<09:36,  2.21s/it, lr: 1.0e-04 loss: 9.107e-02]
anha:  92%|#########2| 3039/3300 [36:58<09:36,  2.21s/it, lr: 1.0e-04 loss: 9.107e-02]
anha:  92%|#########2| 3040/3300 [37:01<09:34,  2.21s/it, lr: 1.0e-04 loss: 1.344e-01]
anha:  92%|#########2| 3040/3300 [37:01<09:34,  2.21s/it, lr: 1.0e-04 loss: 1.344e-01]
anha:  92%|#########2| 3041/3300 [37:01<08:30,  1.97s/it, lr: 1.0e-04 loss: 1.344e-01]
anha:  92%|#########2| 3041/3300 [37:01<08:30,  1.97s/it, lr: 1.0e-04 loss: 1.344e-01]
anha:  92%|#########2| 3041/3300 [37:03<08:30,  1.97s/it, lr: 1.0e-04 loss: 1.815e-01]
anha:  92%|#########2| 3041/3300 [37:03<08:30,  1.97s/it, lr: 1.0e-04 loss: 1.815e-01]
anha:  92%|#########2| 3042/3300 [37:03<08:29,  1.97s/it, lr: 1.0e-04 loss: 1.815e-01]
anha:  92%|#########2| 3042/3300 [37:03<08:29,  1.97s/it, lr: 1.0e-04 loss: 1.815e-01]
anha:  92%|#########2| 3042/3300 [37:05<08:29,  1.97s/it, lr: 1.0e-04 loss: 5.054e-02]
anha:  92%|#########2| 3042/3300 [37:05<08:29,  1.97s/it, lr: 1.0e-04 loss: 5.054e-02]
anha:  92%|#########2| 3043/3300 [37:05<08:25,  1.97s/it, lr: 1.0e-04 loss: 5.054e-02]
anha:  92%|#########2| 3043/3300 [37:05<08:25,  1.97s/it, lr: 1.0e-04 loss: 5.054e-02]
anha:  92%|#########2| 3043/3300 [37:06<08:25,  1.97s/it, lr: 1.0e-04 loss: 5.704e-02]
anha:  92%|#########2| 3043/3300 [37:06<08:25,  1.97s/it, lr: 1.0e-04 loss: 5.704e-02]
anha:  92%|#########2| 3044/3300 [37:06<07:17,  1.71s/it, lr: 1.0e-04 loss: 5.704e-02]
anha:  92%|#########2| 3044/3300 [37:06<07:17,  1.71s/it, lr: 1.0e-04 loss: 5.704e-02]
anha:  92%|#########2| 3044/3300 [37:07<07:17,  1.71s/it, lr: 1.0e-04 loss: 3.865e-02]
anha:  92%|#########2| 3044/3300 [37:07<07:17,  1.71s/it, lr: 1.0e-04 loss: 3.865e-02]
anha:  92%|#########2| 3045/3300 [37:07<06:24,  1.51s/it, lr: 1.0e-04 loss: 3.865e-02]
anha:  92%|#########2| 3045/3300 [37:07<06:24,  1.51s/it, lr: 1.0e-04 loss: 3.865e-02]
anha:  92%|#########2| 3045/3300 [37:08<06:24,  1.51s/it, lr: 1.0e-04 loss: 2.207e-01]
anha:  92%|#########2| 3045/3300 [37:08<06:24,  1.51s/it, lr: 1.0e-04 loss: 2.207e-01]
anha:  92%|#########2| 3046/3300 [37:08<05:45,  1.36s/it, lr: 1.0e-04 loss: 2.207e-01]
anha:  92%|#########2| 3046/3300 [37:08<05:45,  1.36s/it, lr: 1.0e-04 loss: 2.207e-01]
anha:  92%|#########2| 3046/3300 [37:09<05:45,  1.36s/it, lr: 1.0e-04 loss: 6.405e-02]
anha:  92%|#########2| 3046/3300 [37:09<05:45,  1.36s/it, lr: 1.0e-04 loss: 6.405e-02]
anha:  92%|#########2| 3047/3300 [37:09<05:25,  1.29s/it, lr: 1.0e-04 loss: 6.405e-02]
anha:  92%|#########2| 3047/3300 [37:09<05:25,  1.29s/it, lr: 1.0e-04 loss: 6.405e-02]
anha:  92%|#########2| 3047/3300 [37:10<05:25,  1.29s/it, lr: 1.0e-04 loss: 6.855e-02]
anha:  92%|#########2| 3047/3300 [37:10<05:25,  1.29s/it, lr: 1.0e-04 loss: 6.855e-02]
anha:  92%|#########2| 3048/3300 [37:10<05:02,  1.20s/it, lr: 1.0e-04 loss: 6.855e-02]
anha:  92%|#########2| 3048/3300 [37:10<05:02,  1.20s/it, lr: 1.0e-04 loss: 6.855e-02]
anha:  92%|#########2| 3048/3300 [37:12<05:02,  1.20s/it, lr: 1.0e-04 loss: 4.603e-02]
anha:  92%|#########2| 3048/3300 [37:12<05:02,  1.20s/it, lr: 1.0e-04 loss: 4.603e-02]
anha:  92%|#########2| 3049/3300 [37:12<05:40,  1.36s/it, lr: 1.0e-04 loss: 4.603e-02]
anha:  92%|#########2| 3049/3300 [37:12<05:40,  1.36s/it, lr: 1.0e-04 loss: 4.603e-02]
anha:  92%|#########2| 3049/3300 [37:14<05:40,  1.36s/it, lr: 1.0e-04 loss: 4.975e-02]
anha:  92%|#########2| 3049/3300 [37:14<05:40,  1.36s/it, lr: 1.0e-04 loss: 4.975e-02]
anha:  92%|#########2| 3050/3300 [37:15<05:39,  1.36s/it, lr: 1.0e-04 loss: 3.328e-02]
anha:  92%|#########2| 3050/3300 [37:15<05:39,  1.36s/it, lr: 1.0e-04 loss: 3.328e-02]
anha:  92%|#########2| 3051/3300 [37:15<04:00,  1.04it/s, lr: 1.0e-04 loss: 3.328e-02]
anha:  92%|#########2| 3051/3300 [37:15<04:00,  1.04it/s, lr: 1.0e-04 loss: 3.328e-02]
anha:  92%|#########2| 3051/3300 [37:18<04:00,  1.04it/s, lr: 1.0e-04 loss: 1.155e-01]
anha:  92%|#########2| 3051/3300 [37:18<04:00,  1.04it/s, lr: 1.0e-04 loss: 1.155e-01]
anha:  92%|#########2| 3052/3300 [37:18<06:26,  1.56s/it, lr: 1.0e-04 loss: 1.155e-01]
anha:  92%|#########2| 3052/3300 [37:18<06:26,  1.56s/it, lr: 1.0e-04 loss: 1.155e-01]
anha:  92%|#########2| 3052/3300 [37:21<06:26,  1.56s/it, lr: 1.0e-04 loss: 3.465e-02]
anha:  92%|#########2| 3052/3300 [37:21<06:26,  1.56s/it, lr: 1.0e-04 loss: 3.465e-02]
anha:  93%|#########2| 3053/3300 [37:21<07:27,  1.81s/it, lr: 1.0e-04 loss: 3.465e-02]
anha:  93%|#########2| 3053/3300 [37:21<07:27,  1.81s/it, lr: 1.0e-04 loss: 3.465e-02]
anha:  93%|#########2| 3053/3300 [37:23<07:27,  1.81s/it, lr: 1.0e-04 loss: 3.039e-01]
anha:  93%|#########2| 3053/3300 [37:23<07:27,  1.81s/it, lr: 1.0e-04 loss: 3.039e-01]
anha:  93%|#########2| 3054/3300 [37:23<07:36,  1.86s/it, lr: 1.0e-04 loss: 3.039e-01]
anha:  93%|#########2| 3054/3300 [37:23<07:36,  1.86s/it, lr: 1.0e-04 loss: 3.039e-01]
anha:  93%|#########2| 3054/3300 [37:26<07:36,  1.86s/it, lr: 1.0e-04 loss: 4.287e-02]
anha:  93%|#########2| 3054/3300 [37:26<07:36,  1.86s/it, lr: 1.0e-04 loss: 4.287e-02]
anha:  93%|#########2| 3055/3300 [37:26<09:25,  2.31s/it, lr: 1.0e-04 loss: 4.287e-02]
anha:  93%|#########2| 3055/3300 [37:26<09:25,  2.31s/it, lr: 1.0e-04 loss: 4.287e-02]
anha:  93%|#########2| 3055/3300 [37:28<09:25,  2.31s/it, lr: 1.0e-04 loss: 1.403e-01]
anha:  93%|#########2| 3055/3300 [37:28<09:25,  2.31s/it, lr: 1.0e-04 loss: 1.403e-01]
anha:  93%|#########2| 3056/3300 [37:28<08:59,  2.21s/it, lr: 1.0e-04 loss: 1.403e-01]
anha:  93%|#########2| 3056/3300 [37:28<08:59,  2.21s/it, lr: 1.0e-04 loss: 1.403e-01]
anha:  93%|#########2| 3056/3300 [37:29<08:59,  2.21s/it, lr: 1.0e-04 loss: 5.697e-02]
anha:  93%|#########2| 3056/3300 [37:29<08:59,  2.21s/it, lr: 1.0e-04 loss: 5.697e-02]
anha:  93%|#########2| 3057/3300 [37:29<07:40,  1.90s/it, lr: 1.0e-04 loss: 5.697e-02]
anha:  93%|#########2| 3057/3300 [37:29<07:40,  1.90s/it, lr: 1.0e-04 loss: 5.697e-02]
anha:  93%|#########2| 3057/3300 [37:31<07:40,  1.90s/it, lr: 1.0e-04 loss: 3.548e-02]
anha:  93%|#########2| 3057/3300 [37:31<07:40,  1.90s/it, lr: 1.0e-04 loss: 3.548e-02]
anha:  93%|#########2| 3058/3300 [37:31<07:43,  1.92s/it, lr: 1.0e-04 loss: 3.548e-02]
anha:  93%|#########2| 3058/3300 [37:31<07:43,  1.92s/it, lr: 1.0e-04 loss: 3.548e-02]
anha:  93%|#########2| 3058/3300 [37:32<07:43,  1.92s/it, lr: 1.0e-04 loss: 4.618e-02]
anha:  93%|#########2| 3058/3300 [37:32<07:43,  1.92s/it, lr: 1.0e-04 loss: 4.618e-02]
anha:  93%|#########2| 3059/3300 [37:32<06:38,  1.65s/it, lr: 1.0e-04 loss: 4.618e-02]
anha:  93%|#########2| 3059/3300 [37:32<06:38,  1.65s/it, lr: 1.0e-04 loss: 4.618e-02]
anha:  93%|#########2| 3059/3300 [37:36<06:38,  1.65s/it, lr: 1.0e-04 loss: 2.844e-02]
anha:  93%|#########2| 3059/3300 [37:36<06:38,  1.65s/it, lr: 1.0e-04 loss: 2.844e-02]
anha:  93%|#########2| 3060/3300 [37:39<06:36,  1.65s/it, lr: 1.0e-04 loss: 1.040e-01]
anha:  93%|#########2| 3060/3300 [37:39<06:36,  1.65s/it, lr: 1.0e-04 loss: 1.040e-01]
anha:  93%|#########2| 3061/3300 [37:39<06:38,  1.67s/it, lr: 1.0e-04 loss: 1.040e-01]
anha:  93%|#########2| 3061/3300 [37:39<06:38,  1.67s/it, lr: 1.0e-04 loss: 1.040e-01]
anha:  93%|#########2| 3061/3300 [37:41<06:38,  1.67s/it, lr: 1.0e-04 loss: 1.784e-01]
anha:  93%|#########2| 3061/3300 [37:41<06:38,  1.67s/it, lr: 1.0e-04 loss: 1.784e-01]
anha:  93%|#########2| 3062/3300 [37:41<06:40,  1.68s/it, lr: 1.0e-04 loss: 1.784e-01]
anha:  93%|#########2| 3062/3300 [37:41<06:40,  1.68s/it, lr: 1.0e-04 loss: 1.784e-01]
anha:  93%|#########2| 3062/3300 [37:42<06:40,  1.68s/it, lr: 1.0e-04 loss: 5.486e-02]
anha:  93%|#########2| 3062/3300 [37:42<06:40,  1.68s/it, lr: 1.0e-04 loss: 5.486e-02]
anha:  93%|#########2| 3063/3300 [37:42<05:56,  1.51s/it, lr: 1.0e-04 loss: 5.486e-02]
anha:  93%|#########2| 3063/3300 [37:42<05:56,  1.51s/it, lr: 1.0e-04 loss: 5.486e-02]
anha:  93%|#########2| 3063/3300 [37:44<05:56,  1.51s/it, lr: 1.0e-04 loss: 5.037e-02]
anha:  93%|#########2| 3063/3300 [37:44<05:56,  1.51s/it, lr: 1.0e-04 loss: 5.037e-02]
anha:  93%|#########2| 3064/3300 [37:44<06:09,  1.57s/it, lr: 1.0e-04 loss: 5.037e-02]
anha:  93%|#########2| 3064/3300 [37:44<06:09,  1.57s/it, lr: 1.0e-04 loss: 5.037e-02]
anha:  93%|#########2| 3064/3300 [37:46<06:09,  1.57s/it, lr: 1.0e-04 loss: 3.799e-02]
anha:  93%|#########2| 3064/3300 [37:46<06:09,  1.57s/it, lr: 1.0e-04 loss: 3.799e-02]
anha:  93%|#########2| 3065/3300 [37:46<06:33,  1.67s/it, lr: 1.0e-04 loss: 3.799e-02]
anha:  93%|#########2| 3065/3300 [37:46<06:33,  1.67s/it, lr: 1.0e-04 loss: 3.799e-02]
anha:  93%|#########2| 3065/3300 [37:49<06:33,  1.67s/it, lr: 1.0e-04 loss: 1.593e-01]
anha:  93%|#########2| 3065/3300 [37:49<06:33,  1.67s/it, lr: 1.0e-04 loss: 1.593e-01]
anha:  93%|#########2| 3066/3300 [37:49<08:18,  2.13s/it, lr: 1.0e-04 loss: 1.593e-01]
anha:  93%|#########2| 3066/3300 [37:49<08:18,  2.13s/it, lr: 1.0e-04 loss: 1.593e-01]
anha:  93%|#########2| 3066/3300 [37:51<08:18,  2.13s/it, lr: 1.0e-04 loss: 5.043e-02]
anha:  93%|#########2| 3066/3300 [37:51<08:18,  2.13s/it, lr: 1.0e-04 loss: 5.043e-02]
anha:  93%|#########2| 3067/3300 [37:51<07:49,  2.02s/it, lr: 1.0e-04 loss: 5.043e-02]
anha:  93%|#########2| 3067/3300 [37:51<07:49,  2.02s/it, lr: 1.0e-04 loss: 5.043e-02]
anha:  93%|#########2| 3067/3300 [37:52<07:49,  2.02s/it, lr: 1.0e-04 loss: 1.419e-01]
anha:  93%|#########2| 3067/3300 [37:52<07:49,  2.02s/it, lr: 1.0e-04 loss: 1.419e-01]
anha:  93%|#########2| 3068/3300 [37:52<07:29,  1.94s/it, lr: 1.0e-04 loss: 1.419e-01]
anha:  93%|#########2| 3068/3300 [37:52<07:29,  1.94s/it, lr: 1.0e-04 loss: 1.419e-01]
anha:  93%|#########2| 3068/3300 [37:53<07:29,  1.94s/it, lr: 1.0e-04 loss: 1.546e-01]
anha:  93%|#########2| 3068/3300 [37:53<07:29,  1.94s/it, lr: 1.0e-04 loss: 1.546e-01]
anha:  93%|#########3| 3069/3300 [37:53<06:24,  1.66s/it, lr: 1.0e-04 loss: 1.546e-01]
anha:  93%|#########3| 3069/3300 [37:53<06:24,  1.66s/it, lr: 1.0e-04 loss: 1.546e-01]
anha:  93%|#########3| 3069/3300 [37:57<06:24,  1.66s/it, lr: 1.0e-04 loss: 7.905e-02]
anha:  93%|#########3| 3069/3300 [37:57<06:24,  1.66s/it, lr: 1.0e-04 loss: 7.905e-02]
anha:  93%|#########3| 3070/3300 [37:58<06:22,  1.66s/it, lr: 1.0e-04 loss: 2.470e-01]
anha:  93%|#########3| 3070/3300 [37:58<06:22,  1.66s/it, lr: 1.0e-04 loss: 2.470e-01]
anha:  93%|#########3| 3071/3300 [37:58<04:19,  1.13s/it, lr: 1.0e-04 loss: 2.470e-01]
anha:  93%|#########3| 3071/3300 [37:58<04:19,  1.13s/it, lr: 1.0e-04 loss: 2.470e-01]
anha:  93%|#########3| 3071/3300 [37:59<04:19,  1.13s/it, lr: 1.0e-04 loss: 3.762e-02]
anha:  93%|#########3| 3071/3300 [37:59<04:19,  1.13s/it, lr: 1.0e-04 loss: 3.762e-02]
anha:  93%|#########3| 3072/3300 [37:59<04:11,  1.10s/it, lr: 1.0e-04 loss: 3.762e-02]
anha:  93%|#########3| 3072/3300 [37:59<04:11,  1.10s/it, lr: 1.0e-04 loss: 3.762e-02]
anha:  93%|#########3| 3072/3300 [38:02<04:11,  1.10s/it, lr: 1.0e-04 loss: 3.330e-02]
anha:  93%|#########3| 3072/3300 [38:02<04:11,  1.10s/it, lr: 1.0e-04 loss: 3.330e-02]
anha:  93%|#########3| 3073/3300 [38:02<06:24,  1.69s/it, lr: 1.0e-04 loss: 3.330e-02]
anha:  93%|#########3| 3073/3300 [38:02<06:24,  1.69s/it, lr: 1.0e-04 loss: 3.330e-02]
anha:  93%|#########3| 3073/3300 [38:05<06:24,  1.69s/it, lr: 1.0e-04 loss: 6.067e-02]
anha:  93%|#########3| 3073/3300 [38:05<06:24,  1.69s/it, lr: 1.0e-04 loss: 6.067e-02]
anha:  93%|#########3| 3074/3300 [38:05<08:13,  2.18s/it, lr: 1.0e-04 loss: 6.067e-02]
anha:  93%|#########3| 3074/3300 [38:05<08:13,  2.18s/it, lr: 1.0e-04 loss: 6.067e-02]
anha:  93%|#########3| 3074/3300 [38:06<08:13,  2.18s/it, lr: 1.0e-04 loss: 1.121e-01]
anha:  93%|#########3| 3074/3300 [38:06<08:13,  2.18s/it, lr: 1.0e-04 loss: 1.121e-01]
anha:  93%|#########3| 3075/3300 [38:06<06:57,  1.86s/it, lr: 1.0e-04 loss: 1.121e-01]
anha:  93%|#########3| 3075/3300 [38:06<06:57,  1.86s/it, lr: 1.0e-04 loss: 1.121e-01]
anha:  93%|#########3| 3075/3300 [38:08<06:57,  1.86s/it, lr: 1.0e-04 loss: 3.853e-02]
anha:  93%|#########3| 3075/3300 [38:08<06:57,  1.86s/it, lr: 1.0e-04 loss: 3.853e-02]
anha:  93%|#########3| 3076/3300 [38:08<06:08,  1.65s/it, lr: 1.0e-04 loss: 3.853e-02]
anha:  93%|#########3| 3076/3300 [38:08<06:08,  1.65s/it, lr: 1.0e-04 loss: 3.853e-02]
anha:  93%|#########3| 3076/3300 [38:10<06:08,  1.65s/it, lr: 1.0e-04 loss: 1.831e-01]
anha:  93%|#########3| 3076/3300 [38:10<06:08,  1.65s/it, lr: 1.0e-04 loss: 1.831e-01]
anha:  93%|#########3| 3077/3300 [38:10<06:28,  1.74s/it, lr: 1.0e-04 loss: 1.831e-01]
anha:  93%|#########3| 3077/3300 [38:10<06:28,  1.74s/it, lr: 1.0e-04 loss: 1.831e-01]
anha:  93%|#########3| 3077/3300 [38:13<06:28,  1.74s/it, lr: 1.0e-04 loss: 2.333e-02]
anha:  93%|#########3| 3077/3300 [38:13<06:28,  1.74s/it, lr: 1.0e-04 loss: 2.333e-02]
anha:  93%|#########3| 3078/3300 [38:13<08:12,  2.22s/it, lr: 1.0e-04 loss: 2.333e-02]
anha:  93%|#########3| 3078/3300 [38:13<08:12,  2.22s/it, lr: 1.0e-04 loss: 2.333e-02]
anha:  93%|#########3| 3078/3300 [38:15<08:12,  2.22s/it, lr: 1.0e-04 loss: 1.154e-01]
anha:  93%|#########3| 3078/3300 [38:15<08:12,  2.22s/it, lr: 1.0e-04 loss: 1.154e-01]
anha:  93%|#########3| 3079/3300 [38:15<07:54,  2.15s/it, lr: 1.0e-04 loss: 1.154e-01]
anha:  93%|#########3| 3079/3300 [38:15<07:54,  2.15s/it, lr: 1.0e-04 loss: 1.154e-01]
anha:  93%|#########3| 3079/3300 [38:18<07:54,  2.15s/it, lr: 1.0e-04 loss: 1.186e-01]
anha:  93%|#########3| 3079/3300 [38:18<07:54,  2.15s/it, lr: 1.0e-04 loss: 1.186e-01]
anha:  93%|#########3| 3080/3300 [38:19<07:52,  2.15s/it, lr: 1.0e-04 loss: 1.245e-01]
anha:  93%|#########3| 3080/3300 [38:19<07:52,  2.15s/it, lr: 1.0e-04 loss: 1.245e-01]
anha:  93%|#########3| 3081/3300 [38:19<05:05,  1.40s/it, lr: 1.0e-04 loss: 1.245e-01]
anha:  93%|#########3| 3081/3300 [38:19<05:05,  1.40s/it, lr: 1.0e-04 loss: 1.245e-01]
anha:  93%|#########3| 3081/3300 [38:20<05:05,  1.40s/it, lr: 1.0e-04 loss: 5.623e-02]
anha:  93%|#########3| 3081/3300 [38:20<05:05,  1.40s/it, lr: 1.0e-04 loss: 5.623e-02]
anha:  93%|#########3| 3082/3300 [38:20<04:43,  1.30s/it, lr: 1.0e-04 loss: 5.623e-02]
anha:  93%|#########3| 3082/3300 [38:20<04:43,  1.30s/it, lr: 1.0e-04 loss: 5.623e-02]
anha:  93%|#########3| 3082/3300 [38:21<04:43,  1.30s/it, lr: 1.0e-04 loss: 7.707e-02]
anha:  93%|#########3| 3082/3300 [38:21<04:43,  1.30s/it, lr: 1.0e-04 loss: 7.707e-02]
anha:  93%|#########3| 3083/3300 [38:21<04:25,  1.22s/it, lr: 1.0e-04 loss: 7.707e-02]
anha:  93%|#########3| 3083/3300 [38:21<04:25,  1.22s/it, lr: 1.0e-04 loss: 7.707e-02]
anha:  93%|#########3| 3083/3300 [38:25<04:25,  1.22s/it, lr: 1.0e-04 loss: 8.558e-02]
anha:  93%|#########3| 3083/3300 [38:25<04:25,  1.22s/it, lr: 1.0e-04 loss: 8.558e-02]
anha:  93%|#########3| 3084/3300 [38:25<06:33,  1.82s/it, lr: 1.0e-04 loss: 8.558e-02]
anha:  93%|#########3| 3084/3300 [38:25<06:33,  1.82s/it, lr: 1.0e-04 loss: 8.558e-02]
anha:  93%|#########3| 3084/3300 [38:28<06:33,  1.82s/it, lr: 1.0e-04 loss: 1.191e-01]
anha:  93%|#########3| 3084/3300 [38:28<06:33,  1.82s/it, lr: 1.0e-04 loss: 1.191e-01]
anha:  93%|#########3| 3085/3300 [38:28<08:14,  2.30s/it, lr: 1.0e-04 loss: 1.191e-01]
anha:  93%|#########3| 3085/3300 [38:28<08:14,  2.30s/it, lr: 1.0e-04 loss: 1.191e-01]
anha:  93%|#########3| 3085/3300 [38:30<08:14,  2.30s/it, lr: 1.0e-04 loss: 3.687e-02]
anha:  93%|#########3| 3085/3300 [38:30<08:14,  2.30s/it, lr: 1.0e-04 loss: 3.687e-02]
anha:  94%|#########3| 3086/3300 [38:30<07:51,  2.20s/it, lr: 1.0e-04 loss: 3.687e-02]
anha:  94%|#########3| 3086/3300 [38:30<07:51,  2.20s/it, lr: 1.0e-04 loss: 3.687e-02]
anha:  94%|#########3| 3086/3300 [38:32<07:51,  2.20s/it, lr: 1.0e-04 loss: 8.413e-02]
anha:  94%|#########3| 3086/3300 [38:32<07:51,  2.20s/it, lr: 1.0e-04 loss: 8.413e-02]
anha:  94%|#########3| 3087/3300 [38:32<07:35,  2.14s/it, lr: 1.0e-04 loss: 8.413e-02]
anha:  94%|#########3| 3087/3300 [38:32<07:35,  2.14s/it, lr: 1.0e-04 loss: 8.413e-02]
anha:  94%|#########3| 3087/3300 [38:34<07:35,  2.14s/it, lr: 1.0e-04 loss: 8.724e-02]
anha:  94%|#########3| 3087/3300 [38:34<07:35,  2.14s/it, lr: 1.0e-04 loss: 8.724e-02]
anha:  94%|#########3| 3088/3300 [38:34<07:07,  2.01s/it, lr: 1.0e-04 loss: 8.724e-02]
anha:  94%|#########3| 3088/3300 [38:34<07:07,  2.01s/it, lr: 1.0e-04 loss: 8.724e-02]
anha:  94%|#########3| 3088/3300 [38:37<07:07,  2.01s/it, lr: 1.0e-04 loss: 7.406e-02]
anha:  94%|#########3| 3088/3300 [38:37<07:07,  2.01s/it, lr: 1.0e-04 loss: 7.406e-02]
anha:  94%|#########3| 3089/3300 [38:37<08:21,  2.37s/it, lr: 1.0e-04 loss: 7.406e-02]
anha:  94%|#########3| 3089/3300 [38:37<08:21,  2.37s/it, lr: 1.0e-04 loss: 7.406e-02]
anha:  94%|#########3| 3089/3300 [38:39<08:21,  2.37s/it, lr: 1.0e-04 loss: 2.793e-02]
anha:  94%|#########3| 3089/3300 [38:39<08:21,  2.37s/it, lr: 1.0e-04 loss: 2.793e-02]
anha:  94%|#########3| 3090/3300 [38:42<08:18,  2.37s/it, lr: 1.0e-04 loss: 3.838e-02]
anha:  94%|#########3| 3090/3300 [38:42<08:18,  2.37s/it, lr: 1.0e-04 loss: 3.838e-02]
anha:  94%|#########3| 3091/3300 [38:42<07:15,  2.08s/it, lr: 1.0e-04 loss: 3.838e-02]
anha:  94%|#########3| 3091/3300 [38:42<07:15,  2.08s/it, lr: 1.0e-04 loss: 3.838e-02]
anha:  94%|#########3| 3091/3300 [38:44<07:15,  2.08s/it, lr: 1.0e-04 loss: 4.313e-02]
anha:  94%|#########3| 3091/3300 [38:44<07:15,  2.08s/it, lr: 1.0e-04 loss: 4.313e-02]
anha:  94%|#########3| 3092/3300 [38:44<07:08,  2.06s/it, lr: 1.0e-04 loss: 4.313e-02]
anha:  94%|#########3| 3092/3300 [38:44<07:08,  2.06s/it, lr: 1.0e-04 loss: 4.313e-02]
anha:  94%|#########3| 3092/3300 [38:46<07:08,  2.06s/it, lr: 1.0e-04 loss: 5.126e-01]
anha:  94%|#########3| 3092/3300 [38:46<07:08,  2.06s/it, lr: 1.0e-04 loss: 5.126e-01]
anha:  94%|#########3| 3093/3300 [38:46<07:01,  2.04s/it, lr: 1.0e-04 loss: 5.126e-01]
anha:  94%|#########3| 3093/3300 [38:46<07:01,  2.04s/it, lr: 1.0e-04 loss: 5.126e-01]
anha:  94%|#########3| 3093/3300 [38:48<07:01,  2.04s/it, lr: 1.0e-04 loss: 5.448e-02]
anha:  94%|#########3| 3093/3300 [38:48<07:01,  2.04s/it, lr: 1.0e-04 loss: 5.448e-02]
anha:  94%|#########3| 3094/3300 [38:48<06:54,  2.01s/it, lr: 1.0e-04 loss: 5.448e-02]
anha:  94%|#########3| 3094/3300 [38:48<06:54,  2.01s/it, lr: 1.0e-04 loss: 5.448e-02]
anha:  94%|#########3| 3094/3300 [38:49<06:54,  2.01s/it, lr: 1.0e-04 loss: 6.309e-02]
anha:  94%|#########3| 3094/3300 [38:49<06:54,  2.01s/it, lr: 1.0e-04 loss: 6.309e-02]
anha:  94%|#########3| 3095/3300 [38:49<06:01,  1.76s/it, lr: 1.0e-04 loss: 6.309e-02]
anha:  94%|#########3| 3095/3300 [38:49<06:01,  1.76s/it, lr: 1.0e-04 loss: 6.309e-02]
anha:  94%|#########3| 3095/3300 [38:50<06:01,  1.76s/it, lr: 1.0e-04 loss: 5.827e-02]
anha:  94%|#########3| 3095/3300 [38:50<06:01,  1.76s/it, lr: 1.0e-04 loss: 5.827e-02]
anha:  94%|#########3| 3096/3300 [38:50<05:15,  1.55s/it, lr: 1.0e-04 loss: 5.827e-02]
anha:  94%|#########3| 3096/3300 [38:50<05:15,  1.55s/it, lr: 1.0e-04 loss: 5.827e-02]
anha:  94%|#########3| 3096/3300 [38:51<05:15,  1.55s/it, lr: 1.0e-04 loss: 1.015e-01]
anha:  94%|#########3| 3096/3300 [38:51<05:15,  1.55s/it, lr: 1.0e-04 loss: 1.015e-01]
anha:  94%|#########3| 3097/3300 [38:51<04:41,  1.39s/it, lr: 1.0e-04 loss: 1.015e-01]
anha:  94%|#########3| 3097/3300 [38:51<04:41,  1.39s/it, lr: 1.0e-04 loss: 1.015e-01]
anha:  94%|#########3| 3097/3300 [38:52<04:41,  1.39s/it, lr: 1.0e-04 loss: 5.930e-02]
anha:  94%|#########3| 3097/3300 [38:52<04:41,  1.39s/it, lr: 1.0e-04 loss: 5.930e-02]
anha:  94%|#########3| 3098/3300 [38:52<04:17,  1.27s/it, lr: 1.0e-04 loss: 5.930e-02]
anha:  94%|#########3| 3098/3300 [38:52<04:17,  1.27s/it, lr: 1.0e-04 loss: 5.930e-02]
anha:  94%|#########3| 3098/3300 [38:56<04:17,  1.27s/it, lr: 1.0e-04 loss: 7.218e-02]
anha:  94%|#########3| 3098/3300 [38:56<04:17,  1.27s/it, lr: 1.0e-04 loss: 7.218e-02]
anha:  94%|#########3| 3099/3300 [38:56<06:15,  1.87s/it, lr: 1.0e-04 loss: 7.218e-02]
anha:  94%|#########3| 3099/3300 [38:56<06:15,  1.87s/it, lr: 1.0e-04 loss: 7.218e-02]
anha:  94%|#########3| 3099/3300 [38:59<06:15,  1.87s/it, lr: 1.0e-04 loss: 1.420e-01]
anha:  94%|#########3| 3099/3300 [38:59<06:15,  1.87s/it, lr: 1.0e-04 loss: 1.420e-01]
anha:  94%|#########3| 3100/3300 [39:03<06:13,  1.87s/it, lr: 1.0e-04 loss: 1.120e-01]
anha:  94%|#########3| 3100/3300 [39:03<06:13,  1.87s/it, lr: 1.0e-04 loss: 1.120e-01]
anha:  94%|#########3| 3101/3300 [39:03<05:57,  1.80s/it, lr: 1.0e-04 loss: 1.120e-01]
anha:  94%|#########3| 3101/3300 [39:03<05:57,  1.80s/it, lr: 1.0e-04 loss: 1.120e-01]
anha:  94%|#########3| 3101/3300 [39:05<05:57,  1.80s/it, lr: 1.0e-04 loss: 4.348e-02]
anha:  94%|#########3| 3101/3300 [39:05<05:57,  1.80s/it, lr: 1.0e-04 loss: 4.348e-02]
anha:  94%|#########3| 3102/3300 [39:05<06:04,  1.84s/it, lr: 1.0e-04 loss: 4.348e-02]
anha:  94%|#########3| 3102/3300 [39:05<06:04,  1.84s/it, lr: 1.0e-04 loss: 4.348e-02]
anha:  94%|#########3| 3102/3300 [39:06<06:04,  1.84s/it, lr: 1.0e-04 loss: 1.819e-01]
anha:  94%|#########3| 3102/3300 [39:06<06:04,  1.84s/it, lr: 1.0e-04 loss: 1.819e-01]
anha:  94%|#########4| 3103/3300 [39:06<05:56,  1.81s/it, lr: 1.0e-04 loss: 1.819e-01]
anha:  94%|#########4| 3103/3300 [39:06<05:56,  1.81s/it, lr: 1.0e-04 loss: 1.819e-01]
anha:  94%|#########4| 3103/3300 [39:07<05:56,  1.81s/it, lr: 1.0e-04 loss: 4.084e-02]
anha:  94%|#########4| 3103/3300 [39:07<05:56,  1.81s/it, lr: 1.0e-04 loss: 4.084e-02]
anha:  94%|#########4| 3104/3300 [39:07<05:20,  1.64s/it, lr: 1.0e-04 loss: 4.084e-02]
anha:  94%|#########4| 3104/3300 [39:07<05:20,  1.64s/it, lr: 1.0e-04 loss: 4.084e-02]
anha:  94%|#########4| 3104/3300 [39:09<05:20,  1.64s/it, lr: 1.0e-04 loss: 1.022e-01]
anha:  94%|#########4| 3104/3300 [39:09<05:20,  1.64s/it, lr: 1.0e-04 loss: 1.022e-01]
anha:  94%|#########4| 3105/3300 [39:09<05:24,  1.66s/it, lr: 1.0e-04 loss: 1.022e-01]
anha:  94%|#########4| 3105/3300 [39:09<05:24,  1.66s/it, lr: 1.0e-04 loss: 1.022e-01]
anha:  94%|#########4| 3105/3300 [39:13<05:24,  1.66s/it, lr: 1.0e-04 loss: 4.608e-02]
anha:  94%|#########4| 3105/3300 [39:13<05:24,  1.66s/it, lr: 1.0e-04 loss: 4.608e-02]
anha:  94%|#########4| 3106/3300 [39:13<06:57,  2.15s/it, lr: 1.0e-04 loss: 4.608e-02]
anha:  94%|#########4| 3106/3300 [39:13<06:57,  2.15s/it, lr: 1.0e-04 loss: 4.608e-02]
anha:  94%|#########4| 3106/3300 [39:16<06:57,  2.15s/it, lr: 1.0e-04 loss: 5.907e-02]
anha:  94%|#########4| 3106/3300 [39:16<06:57,  2.15s/it, lr: 1.0e-04 loss: 5.907e-02]
anha:  94%|#########4| 3107/3300 [39:16<07:58,  2.48s/it, lr: 1.0e-04 loss: 5.907e-02]
anha:  94%|#########4| 3107/3300 [39:16<07:58,  2.48s/it, lr: 1.0e-04 loss: 5.907e-02]
anha:  94%|#########4| 3107/3300 [39:19<07:58,  2.48s/it, lr: 1.0e-04 loss: 4.175e-02]
anha:  94%|#########4| 3107/3300 [39:19<07:58,  2.48s/it, lr: 1.0e-04 loss: 4.175e-02]
anha:  94%|#########4| 3108/3300 [39:19<08:46,  2.74s/it, lr: 1.0e-04 loss: 4.175e-02]
anha:  94%|#########4| 3108/3300 [39:19<08:46,  2.74s/it, lr: 1.0e-04 loss: 4.175e-02]
anha:  94%|#########4| 3108/3300 [39:22<08:46,  2.74s/it, lr: 1.0e-04 loss: 3.375e-02]
anha:  94%|#########4| 3108/3300 [39:22<08:46,  2.74s/it, lr: 1.0e-04 loss: 3.375e-02]
anha:  94%|#########4| 3109/3300 [39:22<09:11,  2.89s/it, lr: 1.0e-04 loss: 3.375e-02]
anha:  94%|#########4| 3109/3300 [39:22<09:11,  2.89s/it, lr: 1.0e-04 loss: 3.375e-02]
anha:  94%|#########4| 3109/3300 [39:24<09:11,  2.89s/it, lr: 1.0e-04 loss: 3.413e-02]
anha:  94%|#########4| 3109/3300 [39:24<09:11,  2.89s/it, lr: 1.0e-04 loss: 3.413e-02]
anha:  94%|#########4| 3110/3300 [39:28<09:08,  2.89s/it, lr: 1.0e-04 loss: 1.585e-01]
anha:  94%|#########4| 3110/3300 [39:28<09:08,  2.89s/it, lr: 1.0e-04 loss: 1.585e-01]
anha:  94%|#########4| 3111/3300 [39:28<07:26,  2.36s/it, lr: 1.0e-04 loss: 1.585e-01]
anha:  94%|#########4| 3111/3300 [39:28<07:26,  2.36s/it, lr: 1.0e-04 loss: 1.585e-01]
anha:  94%|#########4| 3111/3300 [39:30<07:26,  2.36s/it, lr: 1.0e-04 loss: 3.025e-02]
anha:  94%|#########4| 3111/3300 [39:30<07:26,  2.36s/it, lr: 1.0e-04 loss: 3.025e-02]
anha:  94%|#########4| 3112/3300 [39:30<07:07,  2.27s/it, lr: 1.0e-04 loss: 3.025e-02]
anha:  94%|#########4| 3112/3300 [39:30<07:07,  2.27s/it, lr: 1.0e-04 loss: 3.025e-02]
anha:  94%|#########4| 3112/3300 [39:33<07:07,  2.27s/it, lr: 1.0e-04 loss: 4.597e-02]
anha:  94%|#########4| 3112/3300 [39:33<07:07,  2.27s/it, lr: 1.0e-04 loss: 4.597e-02]
anha:  94%|#########4| 3113/3300 [39:33<07:56,  2.55s/it, lr: 1.0e-04 loss: 4.597e-02]
anha:  94%|#########4| 3113/3300 [39:33<07:56,  2.55s/it, lr: 1.0e-04 loss: 4.597e-02]
anha:  94%|#########4| 3113/3300 [39:36<07:56,  2.55s/it, lr: 1.0e-04 loss: 1.387e-01]
anha:  94%|#########4| 3113/3300 [39:36<07:56,  2.55s/it, lr: 1.0e-04 loss: 1.387e-01]
anha:  94%|#########4| 3114/3300 [39:36<08:24,  2.71s/it, lr: 1.0e-04 loss: 1.387e-01]
anha:  94%|#########4| 3114/3300 [39:36<08:24,  2.71s/it, lr: 1.0e-04 loss: 1.387e-01]
anha:  94%|#########4| 3114/3300 [39:37<08:24,  2.71s/it, lr: 1.0e-04 loss: 2.427e-01]
anha:  94%|#########4| 3114/3300 [39:37<08:24,  2.71s/it, lr: 1.0e-04 loss: 2.427e-01]
anha:  94%|#########4| 3115/3300 [39:37<06:53,  2.23s/it, lr: 1.0e-04 loss: 2.427e-01]
anha:  94%|#########4| 3115/3300 [39:37<06:53,  2.23s/it, lr: 1.0e-04 loss: 2.427e-01]
anha:  94%|#########4| 3115/3300 [39:38<06:53,  2.23s/it, lr: 1.0e-04 loss: 1.209e-01]
anha:  94%|#########4| 3115/3300 [39:38<06:53,  2.23s/it, lr: 1.0e-04 loss: 1.209e-01]
anha:  94%|#########4| 3116/3300 [39:38<05:46,  1.88s/it, lr: 1.0e-04 loss: 1.209e-01]
anha:  94%|#########4| 3116/3300 [39:38<05:46,  1.88s/it, lr: 1.0e-04 loss: 1.209e-01]
anha:  94%|#########4| 3116/3300 [39:42<05:46,  1.88s/it, lr: 1.0e-04 loss: 9.429e-02]
anha:  94%|#########4| 3116/3300 [39:42<05:46,  1.88s/it, lr: 1.0e-04 loss: 9.429e-02]
anha:  94%|#########4| 3117/3300 [39:42<06:56,  2.28s/it, lr: 1.0e-04 loss: 9.429e-02]
anha:  94%|#########4| 3117/3300 [39:42<06:56,  2.28s/it, lr: 1.0e-04 loss: 9.429e-02]
anha:  94%|#########4| 3117/3300 [39:44<06:56,  2.28s/it, lr: 1.0e-04 loss: 4.498e-02]
anha:  94%|#########4| 3117/3300 [39:44<06:56,  2.28s/it, lr: 1.0e-04 loss: 4.498e-02]
anha:  94%|#########4| 3118/3300 [39:44<06:36,  2.18s/it, lr: 1.0e-04 loss: 4.498e-02]
anha:  94%|#########4| 3118/3300 [39:44<06:36,  2.18s/it, lr: 1.0e-04 loss: 4.498e-02]
anha:  94%|#########4| 3118/3300 [39:46<06:36,  2.18s/it, lr: 1.0e-04 loss: 6.665e-02]
anha:  94%|#########4| 3118/3300 [39:46<06:36,  2.18s/it, lr: 1.0e-04 loss: 6.665e-02]
anha:  95%|#########4| 3119/3300 [39:46<06:23,  2.12s/it, lr: 1.0e-04 loss: 6.665e-02]
anha:  95%|#########4| 3119/3300 [39:46<06:23,  2.12s/it, lr: 1.0e-04 loss: 6.665e-02]
anha:  95%|#########4| 3119/3300 [39:48<06:23,  2.12s/it, lr: 1.0e-04 loss: 4.600e-02]
anha:  95%|#########4| 3119/3300 [39:48<06:23,  2.12s/it, lr: 1.0e-04 loss: 4.600e-02]
anha:  95%|#########4| 3120/3300 [39:51<06:21,  2.12s/it, lr: 1.0e-04 loss: 5.735e-02]
anha:  95%|#########4| 3120/3300 [39:51<06:21,  2.12s/it, lr: 1.0e-04 loss: 5.735e-02]
anha:  95%|#########4| 3121/3300 [39:51<05:48,  1.95s/it, lr: 1.0e-04 loss: 5.735e-02]
anha:  95%|#########4| 3121/3300 [39:51<05:48,  1.95s/it, lr: 1.0e-04 loss: 5.735e-02]
anha:  95%|#########4| 3121/3300 [39:52<05:48,  1.95s/it, lr: 1.0e-04 loss: 6.038e-02]
anha:  95%|#########4| 3121/3300 [39:52<05:48,  1.95s/it, lr: 1.0e-04 loss: 6.038e-02]
anha:  95%|#########4| 3122/3300 [39:52<05:10,  1.74s/it, lr: 1.0e-04 loss: 6.038e-02]
anha:  95%|#########4| 3122/3300 [39:52<05:10,  1.74s/it, lr: 1.0e-04 loss: 6.038e-02]
anha:  95%|#########4| 3122/3300 [39:56<05:10,  1.74s/it, lr: 1.0e-04 loss: 7.347e-02]
anha:  95%|#########4| 3122/3300 [39:56<05:10,  1.74s/it, lr: 1.0e-04 loss: 7.347e-02]
anha:  95%|#########4| 3123/3300 [39:56<06:28,  2.19s/it, lr: 1.0e-04 loss: 7.347e-02]
anha:  95%|#########4| 3123/3300 [39:56<06:28,  2.19s/it, lr: 1.0e-04 loss: 7.347e-02]
anha:  95%|#########4| 3123/3300 [39:58<06:28,  2.19s/it, lr: 1.0e-04 loss: 3.698e-02]
anha:  95%|#########4| 3123/3300 [39:58<06:28,  2.19s/it, lr: 1.0e-04 loss: 3.698e-02]
anha:  95%|#########4| 3124/3300 [39:58<06:14,  2.13s/it, lr: 1.0e-04 loss: 3.698e-02]
anha:  95%|#########4| 3124/3300 [39:58<06:14,  2.13s/it, lr: 1.0e-04 loss: 3.698e-02]
anha:  95%|#########4| 3124/3300 [40:00<06:14,  2.13s/it, lr: 1.0e-04 loss: 5.714e-02]
anha:  95%|#########4| 3124/3300 [40:00<06:14,  2.13s/it, lr: 1.0e-04 loss: 5.714e-02]
anha:  95%|#########4| 3125/3300 [40:00<06:06,  2.09s/it, lr: 1.0e-04 loss: 5.714e-02]
anha:  95%|#########4| 3125/3300 [40:00<06:06,  2.09s/it, lr: 1.0e-04 loss: 5.714e-02]
anha:  95%|#########4| 3125/3300 [40:03<06:06,  2.09s/it, lr: 1.0e-04 loss: 4.466e-02]
anha:  95%|#########4| 3125/3300 [40:03<06:06,  2.09s/it, lr: 1.0e-04 loss: 4.466e-02]
anha:  95%|#########4| 3126/3300 [40:03<07:13,  2.49s/it, lr: 1.0e-04 loss: 4.466e-02]
anha:  95%|#########4| 3126/3300 [40:03<07:13,  2.49s/it, lr: 1.0e-04 loss: 4.466e-02]
anha:  95%|#########4| 3126/3300 [40:04<07:13,  2.49s/it, lr: 1.0e-04 loss: 8.164e-02]
anha:  95%|#########4| 3126/3300 [40:04<07:13,  2.49s/it, lr: 1.0e-04 loss: 8.164e-02]
anha:  95%|#########4| 3127/3300 [40:04<05:57,  2.06s/it, lr: 1.0e-04 loss: 8.164e-02]
anha:  95%|#########4| 3127/3300 [40:04<05:57,  2.06s/it, lr: 1.0e-04 loss: 8.164e-02]
anha:  95%|#########4| 3127/3300 [40:06<05:57,  2.06s/it, lr: 1.0e-04 loss: 1.819e-01]
anha:  95%|#########4| 3127/3300 [40:06<05:57,  2.06s/it, lr: 1.0e-04 loss: 1.819e-01]
anha:  95%|#########4| 3128/3300 [40:06<05:49,  2.03s/it, lr: 1.0e-04 loss: 1.819e-01]
anha:  95%|#########4| 3128/3300 [40:06<05:49,  2.03s/it, lr: 1.0e-04 loss: 1.819e-01]
anha:  95%|#########4| 3128/3300 [40:07<05:49,  2.03s/it, lr: 1.0e-04 loss: 5.925e-02]
anha:  95%|#########4| 3128/3300 [40:07<05:49,  2.03s/it, lr: 1.0e-04 loss: 5.925e-02]
anha:  95%|#########4| 3129/3300 [40:07<04:56,  1.73s/it, lr: 1.0e-04 loss: 5.925e-02]
anha:  95%|#########4| 3129/3300 [40:07<04:56,  1.73s/it, lr: 1.0e-04 loss: 5.925e-02]
anha:  95%|#########4| 3129/3300 [40:10<04:56,  1.73s/it, lr: 1.0e-04 loss: 6.303e-02]
anha:  95%|#########4| 3129/3300 [40:10<04:56,  1.73s/it, lr: 1.0e-04 loss: 6.303e-02]
anha:  95%|#########4| 3130/3300 [40:13<04:54,  1.73s/it, lr: 1.0e-04 loss: 2.478e-01]
anha:  95%|#########4| 3130/3300 [40:13<04:54,  1.73s/it, lr: 1.0e-04 loss: 2.478e-01]
anha:  95%|#########4| 3131/3300 [40:13<04:57,  1.76s/it, lr: 1.0e-04 loss: 2.478e-01]
anha:  95%|#########4| 3131/3300 [40:13<04:57,  1.76s/it, lr: 1.0e-04 loss: 2.478e-01]
anha:  95%|#########4| 3131/3300 [40:14<04:57,  1.76s/it, lr: 1.0e-04 loss: 5.178e-02]
anha:  95%|#########4| 3131/3300 [40:14<04:57,  1.76s/it, lr: 1.0e-04 loss: 5.178e-02]
anha:  95%|#########4| 3132/3300 [40:14<04:24,  1.57s/it, lr: 1.0e-04 loss: 5.178e-02]
anha:  95%|#########4| 3132/3300 [40:14<04:24,  1.57s/it, lr: 1.0e-04 loss: 5.178e-02]
anha:  95%|#########4| 3132/3300 [40:16<04:24,  1.57s/it, lr: 1.0e-04 loss: 1.255e-01]
anha:  95%|#########4| 3132/3300 [40:16<04:24,  1.57s/it, lr: 1.0e-04 loss: 1.255e-01]
anha:  95%|#########4| 3133/3300 [40:16<04:29,  1.62s/it, lr: 1.0e-04 loss: 1.255e-01]
anha:  95%|#########4| 3133/3300 [40:16<04:29,  1.62s/it, lr: 1.0e-04 loss: 1.255e-01]
anha:  95%|#########4| 3133/3300 [40:17<04:29,  1.62s/it, lr: 1.0e-04 loss: 4.910e-02]
anha:  95%|#########4| 3133/3300 [40:17<04:29,  1.62s/it, lr: 1.0e-04 loss: 4.910e-02]
anha:  95%|#########4| 3134/3300 [40:17<04:00,  1.45s/it, lr: 1.0e-04 loss: 4.910e-02]
anha:  95%|#########4| 3134/3300 [40:17<04:00,  1.45s/it, lr: 1.0e-04 loss: 4.910e-02]
anha:  95%|#########4| 3134/3300 [40:19<04:00,  1.45s/it, lr: 1.0e-04 loss: 4.329e-02]
anha:  95%|#########4| 3134/3300 [40:19<04:00,  1.45s/it, lr: 1.0e-04 loss: 4.329e-02]
anha:  95%|#########5| 3135/3300 [40:19<04:23,  1.60s/it, lr: 1.0e-04 loss: 4.329e-02]
anha:  95%|#########5| 3135/3300 [40:19<04:23,  1.60s/it, lr: 1.0e-04 loss: 4.329e-02]
anha:  95%|#########5| 3135/3300 [40:20<04:23,  1.60s/it, lr: 1.0e-04 loss: 5.098e-02]
anha:  95%|#########5| 3135/3300 [40:20<04:23,  1.60s/it, lr: 1.0e-04 loss: 5.098e-02]
anha:  95%|#########5| 3136/3300 [40:20<03:53,  1.42s/it, lr: 1.0e-04 loss: 5.098e-02]
anha:  95%|#########5| 3136/3300 [40:20<03:53,  1.42s/it, lr: 1.0e-04 loss: 5.098e-02]
anha:  95%|#########5| 3136/3300 [40:22<03:53,  1.42s/it, lr: 1.0e-04 loss: 8.787e-02]
anha:  95%|#########5| 3136/3300 [40:22<03:53,  1.42s/it, lr: 1.0e-04 loss: 8.787e-02]
anha:  95%|#########5| 3137/3300 [40:22<04:06,  1.51s/it, lr: 1.0e-04 loss: 8.787e-02]
anha:  95%|#########5| 3137/3300 [40:22<04:06,  1.51s/it, lr: 1.0e-04 loss: 8.787e-02]
anha:  95%|#########5| 3137/3300 [40:23<04:06,  1.51s/it, lr: 1.0e-04 loss: 5.019e-02]
anha:  95%|#########5| 3137/3300 [40:23<04:06,  1.51s/it, lr: 1.0e-04 loss: 5.019e-02]
anha:  95%|#########5| 3138/3300 [40:23<03:40,  1.36s/it, lr: 1.0e-04 loss: 5.019e-02]
anha:  95%|#########5| 3138/3300 [40:23<03:40,  1.36s/it, lr: 1.0e-04 loss: 5.019e-02]
anha:  95%|#########5| 3138/3300 [40:25<03:40,  1.36s/it, lr: 1.0e-04 loss: 3.423e-02]
anha:  95%|#########5| 3138/3300 [40:25<03:40,  1.36s/it, lr: 1.0e-04 loss: 3.423e-02]
anha:  95%|#########5| 3139/3300 [40:25<04:08,  1.55s/it, lr: 1.0e-04 loss: 3.423e-02]
anha:  95%|#########5| 3139/3300 [40:25<04:08,  1.55s/it, lr: 1.0e-04 loss: 3.423e-02]
anha:  95%|#########5| 3139/3300 [40:26<04:08,  1.55s/it, lr: 1.0e-04 loss: 6.073e-02]
anha:  95%|#########5| 3139/3300 [40:26<04:08,  1.55s/it, lr: 1.0e-04 loss: 6.073e-02]
anha:  95%|#########5| 3140/3300 [40:27<04:07,  1.55s/it, lr: 1.0e-04 loss: 6.736e-02]
anha:  95%|#########5| 3140/3300 [40:27<04:07,  1.55s/it, lr: 1.0e-04 loss: 6.736e-02]
anha:  95%|#########5| 3141/3300 [40:27<03:19,  1.26s/it, lr: 1.0e-04 loss: 6.736e-02]
anha:  95%|#########5| 3141/3300 [40:27<03:19,  1.26s/it, lr: 1.0e-04 loss: 6.736e-02]
anha:  95%|#########5| 3141/3300 [40:31<03:19,  1.26s/it, lr: 1.0e-04 loss: 3.648e-02]
anha:  95%|#########5| 3141/3300 [40:31<03:19,  1.26s/it, lr: 1.0e-04 loss: 3.648e-02]
anha:  95%|#########5| 3142/3300 [40:31<04:40,  1.78s/it, lr: 1.0e-04 loss: 3.648e-02]
anha:  95%|#########5| 3142/3300 [40:31<04:40,  1.78s/it, lr: 1.0e-04 loss: 3.648e-02]
anha:  95%|#########5| 3142/3300 [40:33<04:40,  1.78s/it, lr: 1.0e-04 loss: 3.688e-01]
anha:  95%|#########5| 3142/3300 [40:33<04:40,  1.78s/it, lr: 1.0e-04 loss: 3.688e-01]
anha:  95%|#########5| 3143/3300 [40:33<04:46,  1.82s/it, lr: 1.0e-04 loss: 3.688e-01]
anha:  95%|#########5| 3143/3300 [40:33<04:46,  1.82s/it, lr: 1.0e-04 loss: 3.688e-01]
anha:  95%|#########5| 3143/3300 [40:35<04:46,  1.82s/it, lr: 1.0e-04 loss: 3.574e-02]
anha:  95%|#########5| 3143/3300 [40:35<04:46,  1.82s/it, lr: 1.0e-04 loss: 3.574e-02]
anha:  95%|#########5| 3144/3300 [40:35<04:50,  1.86s/it, lr: 1.0e-04 loss: 3.574e-02]
anha:  95%|#########5| 3144/3300 [40:35<04:50,  1.86s/it, lr: 1.0e-04 loss: 3.574e-02]
anha:  95%|#########5| 3144/3300 [40:38<04:50,  1.86s/it, lr: 1.0e-04 loss: 2.408e-02]
anha:  95%|#########5| 3144/3300 [40:38<04:50,  1.86s/it, lr: 1.0e-04 loss: 2.408e-02]
anha:  95%|#########5| 3145/3300 [40:38<05:54,  2.29s/it, lr: 1.0e-04 loss: 2.408e-02]
anha:  95%|#########5| 3145/3300 [40:38<05:54,  2.29s/it, lr: 1.0e-04 loss: 2.408e-02]
anha:  95%|#########5| 3145/3300 [40:39<05:54,  2.29s/it, lr: 1.0e-04 loss: 5.404e-02]
anha:  95%|#########5| 3145/3300 [40:39<05:54,  2.29s/it, lr: 1.0e-04 loss: 5.404e-02]
anha:  95%|#########5| 3146/3300 [40:39<04:55,  1.92s/it, lr: 1.0e-04 loss: 5.404e-02]
anha:  95%|#########5| 3146/3300 [40:39<04:55,  1.92s/it, lr: 1.0e-04 loss: 5.404e-02]
anha:  95%|#########5| 3146/3300 [40:43<04:55,  1.92s/it, lr: 1.0e-04 loss: 1.179e-01]
anha:  95%|#########5| 3146/3300 [40:43<04:55,  1.92s/it, lr: 1.0e-04 loss: 1.179e-01]
anha:  95%|#########5| 3147/3300 [40:43<06:02,  2.37s/it, lr: 1.0e-04 loss: 1.179e-01]
anha:  95%|#########5| 3147/3300 [40:43<06:02,  2.37s/it, lr: 1.0e-04 loss: 1.179e-01]
anha:  95%|#########5| 3147/3300 [40:46<06:02,  2.37s/it, lr: 1.0e-04 loss: 7.709e-02]
anha:  95%|#########5| 3147/3300 [40:46<06:02,  2.37s/it, lr: 1.0e-04 loss: 7.709e-02]
anha:  95%|#########5| 3148/3300 [40:46<06:44,  2.66s/it, lr: 1.0e-04 loss: 7.709e-02]
anha:  95%|#########5| 3148/3300 [40:46<06:44,  2.66s/it, lr: 1.0e-04 loss: 7.709e-02]
anha:  95%|#########5| 3148/3300 [40:49<06:44,  2.66s/it, lr: 1.0e-04 loss: 9.672e-02]
anha:  95%|#########5| 3148/3300 [40:49<06:44,  2.66s/it, lr: 1.0e-04 loss: 9.672e-02]
anha:  95%|#########5| 3149/3300 [40:49<07:10,  2.85s/it, lr: 1.0e-04 loss: 9.672e-02]
anha:  95%|#########5| 3149/3300 [40:49<07:10,  2.85s/it, lr: 1.0e-04 loss: 9.672e-02]
anha:  95%|#########5| 3149/3300 [40:50<07:10,  2.85s/it, lr: 1.0e-04 loss: 3.518e-02]
anha:  95%|#########5| 3149/3300 [40:50<07:10,  2.85s/it, lr: 1.0e-04 loss: 3.518e-02]
anha:  95%|#########5| 3150/3300 [40:51<07:08,  2.85s/it, lr: 1.0e-04 loss: 4.926e-02]
anha:  95%|#########5| 3150/3300 [40:51<07:08,  2.85s/it, lr: 1.0e-04 loss: 4.926e-02]
anha:  95%|#########5| 3151/3300 [40:51<04:24,  1.78s/it, lr: 1.0e-04 loss: 4.926e-02]
anha:  95%|#########5| 3151/3300 [40:51<04:24,  1.78s/it, lr: 1.0e-04 loss: 4.926e-02]
anha:  95%|#########5| 3151/3300 [40:52<04:24,  1.78s/it, lr: 1.0e-04 loss: 1.324e-01]
anha:  95%|#########5| 3151/3300 [40:52<04:24,  1.78s/it, lr: 1.0e-04 loss: 1.324e-01]
anha:  96%|#########5| 3152/3300 [40:52<03:54,  1.58s/it, lr: 1.0e-04 loss: 1.324e-01]
anha:  96%|#########5| 3152/3300 [40:52<03:54,  1.58s/it, lr: 1.0e-04 loss: 1.324e-01]
anha:  96%|#########5| 3152/3300 [40:53<03:54,  1.58s/it, lr: 1.0e-04 loss: 7.224e-02]
anha:  96%|#########5| 3152/3300 [40:53<03:54,  1.58s/it, lr: 1.0e-04 loss: 7.224e-02]
anha:  96%|#########5| 3153/3300 [40:53<03:30,  1.43s/it, lr: 1.0e-04 loss: 7.224e-02]
anha:  96%|#########5| 3153/3300 [40:53<03:30,  1.43s/it, lr: 1.0e-04 loss: 7.224e-02]
anha:  96%|#########5| 3153/3300 [40:54<03:30,  1.43s/it, lr: 1.0e-04 loss: 2.754e-02]
anha:  96%|#########5| 3153/3300 [40:54<03:30,  1.43s/it, lr: 1.0e-04 loss: 2.754e-02]
anha:  96%|#########5| 3154/3300 [40:54<03:12,  1.32s/it, lr: 1.0e-04 loss: 2.754e-02]
anha:  96%|#########5| 3154/3300 [40:54<03:12,  1.32s/it, lr: 1.0e-04 loss: 2.754e-02]
anha:  96%|#########5| 3154/3300 [40:56<03:12,  1.32s/it, lr: 1.0e-04 loss: 4.555e-02]
anha:  96%|#########5| 3154/3300 [40:56<03:12,  1.32s/it, lr: 1.0e-04 loss: 4.555e-02]
anha:  96%|#########5| 3155/3300 [40:56<03:36,  1.49s/it, lr: 1.0e-04 loss: 4.555e-02]
anha:  96%|#########5| 3155/3300 [40:56<03:36,  1.49s/it, lr: 1.0e-04 loss: 4.555e-02]
anha:  96%|#########5| 3155/3300 [40:57<03:36,  1.49s/it, lr: 1.0e-04 loss: 6.673e-02]
anha:  96%|#########5| 3155/3300 [40:57<03:36,  1.49s/it, lr: 1.0e-04 loss: 6.673e-02]
anha:  96%|#########5| 3156/3300 [40:57<03:14,  1.35s/it, lr: 1.0e-04 loss: 6.673e-02]
anha:  96%|#########5| 3156/3300 [40:57<03:14,  1.35s/it, lr: 1.0e-04 loss: 6.673e-02]
anha:  96%|#########5| 3156/3300 [40:58<03:14,  1.35s/it, lr: 1.0e-04 loss: 1.751e-01]
anha:  96%|#########5| 3156/3300 [40:58<03:14,  1.35s/it, lr: 1.0e-04 loss: 1.751e-01]
anha:  96%|#########5| 3157/3300 [40:58<02:58,  1.25s/it, lr: 1.0e-04 loss: 1.751e-01]
anha:  96%|#########5| 3157/3300 [40:58<02:58,  1.25s/it, lr: 1.0e-04 loss: 1.751e-01]
anha:  96%|#########5| 3157/3300 [41:02<02:58,  1.25s/it, lr: 1.0e-04 loss: 1.421e-01]
anha:  96%|#########5| 3157/3300 [41:02<02:58,  1.25s/it, lr: 1.0e-04 loss: 1.421e-01]
anha:  96%|#########5| 3158/3300 [41:02<04:30,  1.90s/it, lr: 1.0e-04 loss: 1.421e-01]
anha:  96%|#########5| 3158/3300 [41:02<04:30,  1.90s/it, lr: 1.0e-04 loss: 1.421e-01]
anha:  96%|#########5| 3158/3300 [41:05<04:30,  1.90s/it, lr: 1.0e-04 loss: 5.384e-02]
anha:  96%|#########5| 3158/3300 [41:05<04:30,  1.90s/it, lr: 1.0e-04 loss: 5.384e-02]
anha:  96%|#########5| 3159/3300 [41:05<05:33,  2.36s/it, lr: 1.0e-04 loss: 5.384e-02]
anha:  96%|#########5| 3159/3300 [41:05<05:33,  2.36s/it, lr: 1.0e-04 loss: 5.384e-02]
anha:  96%|#########5| 3159/3300 [41:07<05:33,  2.36s/it, lr: 1.0e-04 loss: 6.386e-02]
anha:  96%|#########5| 3159/3300 [41:07<05:33,  2.36s/it, lr: 1.0e-04 loss: 6.386e-02]
anha:  96%|#########5| 3160/3300 [41:10<05:30,  2.36s/it, lr: 1.0e-04 loss: 1.091e-01]
anha:  96%|#########5| 3160/3300 [41:10<05:30,  2.36s/it, lr: 1.0e-04 loss: 1.091e-01]
anha:  96%|#########5| 3161/3300 [41:10<04:45,  2.05s/it, lr: 1.0e-04 loss: 1.091e-01]
anha:  96%|#########5| 3161/3300 [41:10<04:45,  2.05s/it, lr: 1.0e-04 loss: 1.091e-01]
anha:  96%|#########5| 3161/3300 [41:14<04:45,  2.05s/it, lr: 1.0e-04 loss: 4.550e-02]
anha:  96%|#########5| 3161/3300 [41:14<04:45,  2.05s/it, lr: 1.0e-04 loss: 4.550e-02]
anha:  96%|#########5| 3162/3300 [41:14<05:28,  2.38s/it, lr: 1.0e-04 loss: 4.550e-02]
anha:  96%|#########5| 3162/3300 [41:14<05:28,  2.38s/it, lr: 1.0e-04 loss: 4.550e-02]
anha:  96%|#########5| 3162/3300 [41:15<05:28,  2.38s/it, lr: 1.0e-04 loss: 5.547e-02]
anha:  96%|#########5| 3162/3300 [41:15<05:28,  2.38s/it, lr: 1.0e-04 loss: 5.547e-02]
anha:  96%|#########5| 3163/3300 [41:15<04:37,  2.02s/it, lr: 1.0e-04 loss: 5.547e-02]
anha:  96%|#########5| 3163/3300 [41:15<04:37,  2.02s/it, lr: 1.0e-04 loss: 5.547e-02]
anha:  96%|#########5| 3163/3300 [41:16<04:37,  2.02s/it, lr: 1.0e-04 loss: 6.043e-02]
anha:  96%|#########5| 3163/3300 [41:16<04:37,  2.02s/it, lr: 1.0e-04 loss: 6.043e-02]
anha:  96%|#########5| 3164/3300 [41:16<03:57,  1.74s/it, lr: 1.0e-04 loss: 6.043e-02]
anha:  96%|#########5| 3164/3300 [41:16<03:57,  1.74s/it, lr: 1.0e-04 loss: 6.043e-02]
anha:  96%|#########5| 3164/3300 [41:18<03:57,  1.74s/it, lr: 1.0e-04 loss: 5.357e-02]
anha:  96%|#########5| 3164/3300 [41:18<03:57,  1.74s/it, lr: 1.0e-04 loss: 5.357e-02]
anha:  96%|#########5| 3165/3300 [41:18<04:03,  1.81s/it, lr: 1.0e-04 loss: 5.357e-02]
anha:  96%|#########5| 3165/3300 [41:18<04:03,  1.81s/it, lr: 1.0e-04 loss: 5.357e-02]
anha:  96%|#########5| 3165/3300 [41:19<04:03,  1.81s/it, lr: 1.0e-04 loss: 9.935e-02]
anha:  96%|#########5| 3165/3300 [41:19<04:03,  1.81s/it, lr: 1.0e-04 loss: 9.935e-02]
anha:  96%|#########5| 3166/3300 [41:19<03:31,  1.58s/it, lr: 1.0e-04 loss: 9.935e-02]
anha:  96%|#########5| 3166/3300 [41:19<03:31,  1.58s/it, lr: 1.0e-04 loss: 9.935e-02]
anha:  96%|#########5| 3166/3300 [41:20<03:31,  1.58s/it, lr: 1.0e-04 loss: 3.359e-02]
anha:  96%|#########5| 3166/3300 [41:20<03:31,  1.58s/it, lr: 1.0e-04 loss: 3.359e-02]
anha:  96%|#########5| 3167/3300 [41:20<03:35,  1.62s/it, lr: 1.0e-04 loss: 3.359e-02]
anha:  96%|#########5| 3167/3300 [41:20<03:35,  1.62s/it, lr: 1.0e-04 loss: 3.359e-02]
anha:  96%|#########5| 3167/3300 [41:23<03:35,  1.62s/it, lr: 1.0e-04 loss: 4.731e-02]
anha:  96%|#########5| 3167/3300 [41:23<03:35,  1.62s/it, lr: 1.0e-04 loss: 4.731e-02]
anha:  96%|#########6| 3168/3300 [41:23<03:52,  1.76s/it, lr: 1.0e-04 loss: 4.731e-02]
anha:  96%|#########6| 3168/3300 [41:23<03:52,  1.76s/it, lr: 1.0e-04 loss: 4.731e-02]
anha:  96%|#########6| 3168/3300 [41:24<03:52,  1.76s/it, lr: 1.0e-04 loss: 3.558e-02]
anha:  96%|#########6| 3168/3300 [41:24<03:52,  1.76s/it, lr: 1.0e-04 loss: 3.558e-02]
anha:  96%|#########6| 3169/3300 [41:24<03:21,  1.54s/it, lr: 1.0e-04 loss: 3.558e-02]
anha:  96%|#########6| 3169/3300 [41:24<03:21,  1.54s/it, lr: 1.0e-04 loss: 3.558e-02]
anha:  96%|#########6| 3169/3300 [41:27<03:21,  1.54s/it, lr: 1.0e-04 loss: 4.170e-02]
anha:  96%|#########6| 3169/3300 [41:27<03:21,  1.54s/it, lr: 1.0e-04 loss: 4.170e-02]
anha:  96%|#########6| 3170/3300 [41:30<03:19,  1.54s/it, lr: 1.0e-04 loss: 5.279e-02]
anha:  96%|#########6| 3170/3300 [41:30<03:19,  1.54s/it, lr: 1.0e-04 loss: 5.279e-02]
anha:  96%|#########6| 3171/3300 [41:30<03:28,  1.62s/it, lr: 1.0e-04 loss: 5.279e-02]
anha:  96%|#########6| 3171/3300 [41:30<03:28,  1.62s/it, lr: 1.0e-04 loss: 5.279e-02]
anha:  96%|#########6| 3171/3300 [41:32<03:28,  1.62s/it, lr: 1.0e-04 loss: 5.494e-02]
anha:  96%|#########6| 3171/3300 [41:32<03:28,  1.62s/it, lr: 1.0e-04 loss: 5.494e-02]
anha:  96%|#########6| 3172/3300 [41:32<03:37,  1.70s/it, lr: 1.0e-04 loss: 5.494e-02]
anha:  96%|#########6| 3172/3300 [41:32<03:37,  1.70s/it, lr: 1.0e-04 loss: 5.494e-02]
anha:  96%|#########6| 3172/3300 [41:34<03:37,  1.70s/it, lr: 1.0e-04 loss: 1.984e-01]
anha:  96%|#########6| 3172/3300 [41:34<03:37,  1.70s/it, lr: 1.0e-04 loss: 1.984e-01]
anha:  96%|#########6| 3173/3300 [41:34<03:36,  1.71s/it, lr: 1.0e-04 loss: 1.984e-01]
anha:  96%|#########6| 3173/3300 [41:34<03:36,  1.71s/it, lr: 1.0e-04 loss: 1.984e-01]
anha:  96%|#########6| 3173/3300 [41:36<03:36,  1.71s/it, lr: 1.0e-04 loss: 4.021e-02]
anha:  96%|#########6| 3173/3300 [41:36<03:36,  1.71s/it, lr: 1.0e-04 loss: 4.021e-02]
anha:  96%|#########6| 3174/3300 [41:36<03:44,  1.78s/it, lr: 1.0e-04 loss: 4.021e-02]
anha:  96%|#########6| 3174/3300 [41:36<03:44,  1.78s/it, lr: 1.0e-04 loss: 4.021e-02]
anha:  96%|#########6| 3174/3300 [41:39<03:44,  1.78s/it, lr: 1.0e-04 loss: 4.355e-02]
anha:  96%|#########6| 3174/3300 [41:39<03:44,  1.78s/it, lr: 1.0e-04 loss: 4.355e-02]
anha:  96%|#########6| 3175/3300 [41:39<04:35,  2.21s/it, lr: 1.0e-04 loss: 4.355e-02]
anha:  96%|#########6| 3175/3300 [41:39<04:35,  2.21s/it, lr: 1.0e-04 loss: 4.355e-02]
anha:  96%|#########6| 3175/3300 [41:41<04:35,  2.21s/it, lr: 1.0e-04 loss: 5.439e-02]
anha:  96%|#########6| 3175/3300 [41:41<04:35,  2.21s/it, lr: 1.0e-04 loss: 5.439e-02]
anha:  96%|#########6| 3176/3300 [41:41<04:25,  2.14s/it, lr: 1.0e-04 loss: 5.439e-02]
anha:  96%|#########6| 3176/3300 [41:41<04:25,  2.14s/it, lr: 1.0e-04 loss: 5.439e-02]
anha:  96%|#########6| 3176/3300 [41:45<04:25,  2.14s/it, lr: 1.0e-04 loss: 4.246e-02]
anha:  96%|#########6| 3176/3300 [41:45<04:25,  2.14s/it, lr: 1.0e-04 loss: 4.246e-02]
anha:  96%|#########6| 3177/3300 [41:45<05:04,  2.47s/it, lr: 1.0e-04 loss: 4.246e-02]
anha:  96%|#########6| 3177/3300 [41:45<05:04,  2.47s/it, lr: 1.0e-04 loss: 4.246e-02]
anha:  96%|#########6| 3177/3300 [41:47<05:04,  2.47s/it, lr: 1.0e-04 loss: 5.578e-02]
anha:  96%|#########6| 3177/3300 [41:47<05:04,  2.47s/it, lr: 1.0e-04 loss: 5.578e-02]
anha:  96%|#########6| 3178/3300 [41:47<04:43,  2.32s/it, lr: 1.0e-04 loss: 5.578e-02]
anha:  96%|#########6| 3178/3300 [41:47<04:43,  2.32s/it, lr: 1.0e-04 loss: 5.578e-02]
anha:  96%|#########6| 3178/3300 [41:48<04:43,  2.32s/it, lr: 1.0e-04 loss: 3.758e-02]
anha:  96%|#########6| 3178/3300 [41:48<04:43,  2.32s/it, lr: 1.0e-04 loss: 3.758e-02]
anha:  96%|#########6| 3179/3300 [41:48<03:54,  1.94s/it, lr: 1.0e-04 loss: 3.758e-02]
anha:  96%|#########6| 3179/3300 [41:48<03:54,  1.94s/it, lr: 1.0e-04 loss: 3.758e-02]
anha:  96%|#########6| 3179/3300 [41:49<03:54,  1.94s/it, lr: 1.0e-04 loss: 1.446e-01]
anha:  96%|#########6| 3179/3300 [41:49<03:54,  1.94s/it, lr: 1.0e-04 loss: 1.446e-01]
anha:  96%|#########6| 3180/3300 [41:52<03:52,  1.94s/it, lr: 1.0e-04 loss: 7.428e-02]
anha:  96%|#########6| 3180/3300 [41:52<03:52,  1.94s/it, lr: 1.0e-04 loss: 7.428e-02]
anha:  96%|#########6| 3181/3300 [41:52<03:39,  1.85s/it, lr: 1.0e-04 loss: 7.428e-02]
anha:  96%|#########6| 3181/3300 [41:52<03:39,  1.85s/it, lr: 1.0e-04 loss: 7.428e-02]
anha:  96%|#########6| 3181/3300 [41:55<03:39,  1.85s/it, lr: 1.0e-04 loss: 1.333e-01]
anha:  96%|#########6| 3181/3300 [41:55<03:39,  1.85s/it, lr: 1.0e-04 loss: 1.333e-01]
anha:  96%|#########6| 3182/3300 [41:55<04:19,  2.20s/it, lr: 1.0e-04 loss: 1.333e-01]
anha:  96%|#########6| 3182/3300 [41:55<04:19,  2.20s/it, lr: 1.0e-04 loss: 1.333e-01]
anha:  96%|#########6| 3182/3300 [41:59<04:19,  2.20s/it, lr: 1.0e-04 loss: 7.641e-02]
anha:  96%|#########6| 3182/3300 [41:59<04:19,  2.20s/it, lr: 1.0e-04 loss: 7.641e-02]
anha:  96%|#########6| 3183/3300 [41:59<04:49,  2.47s/it, lr: 1.0e-04 loss: 7.641e-02]
anha:  96%|#########6| 3183/3300 [41:59<04:49,  2.47s/it, lr: 1.0e-04 loss: 7.641e-02]
anha:  96%|#########6| 3183/3300 [42:02<04:49,  2.47s/it, lr: 1.0e-04 loss: 4.903e-02]
anha:  96%|#########6| 3183/3300 [42:02<04:49,  2.47s/it, lr: 1.0e-04 loss: 4.903e-02]
anha:  96%|#########6| 3184/3300 [42:02<05:12,  2.70s/it, lr: 1.0e-04 loss: 4.903e-02]
anha:  96%|#########6| 3184/3300 [42:02<05:12,  2.70s/it, lr: 1.0e-04 loss: 4.903e-02]
anha:  96%|#########6| 3184/3300 [42:05<05:12,  2.70s/it, lr: 1.0e-04 loss: 1.109e-01]
anha:  96%|#########6| 3184/3300 [42:05<05:12,  2.70s/it, lr: 1.0e-04 loss: 1.109e-01]
anha:  97%|#########6| 3185/3300 [42:05<05:35,  2.92s/it, lr: 1.0e-04 loss: 1.109e-01]
anha:  97%|#########6| 3185/3300 [42:05<05:35,  2.92s/it, lr: 1.0e-04 loss: 1.109e-01]
anha:  97%|#########6| 3185/3300 [42:09<05:35,  2.92s/it, lr: 1.0e-04 loss: 1.249e-01]
anha:  97%|#########6| 3185/3300 [42:09<05:35,  2.92s/it, lr: 1.0e-04 loss: 1.249e-01]
anha:  97%|#########6| 3186/3300 [42:09<05:51,  3.08s/it, lr: 1.0e-04 loss: 1.249e-01]
anha:  97%|#########6| 3186/3300 [42:09<05:51,  3.08s/it, lr: 1.0e-04 loss: 1.249e-01]
anha:  97%|#########6| 3186/3300 [42:12<05:51,  3.08s/it, lr: 1.0e-04 loss: 8.258e-02]
anha:  97%|#########6| 3186/3300 [42:12<05:51,  3.08s/it, lr: 1.0e-04 loss: 8.258e-02]
anha:  97%|#########6| 3187/3300 [42:12<06:02,  3.21s/it, lr: 1.0e-04 loss: 8.258e-02]
anha:  97%|#########6| 3187/3300 [42:12<06:02,  3.21s/it, lr: 1.0e-04 loss: 8.258e-02]
anha:  97%|#########6| 3187/3300 [42:16<06:02,  3.21s/it, lr: 1.0e-04 loss: 4.906e-02]
anha:  97%|#########6| 3187/3300 [42:16<06:02,  3.21s/it, lr: 1.0e-04 loss: 4.906e-02]
anha:  97%|#########6| 3188/3300 [42:16<05:58,  3.20s/it, lr: 1.0e-04 loss: 4.906e-02]
anha:  97%|#########6| 3188/3300 [42:16<05:58,  3.20s/it, lr: 1.0e-04 loss: 4.906e-02]
anha:  97%|#########6| 3188/3300 [42:17<05:58,  3.20s/it, lr: 1.0e-04 loss: 5.274e-02]
anha:  97%|#########6| 3188/3300 [42:17<05:58,  3.20s/it, lr: 1.0e-04 loss: 5.274e-02]
anha:  97%|#########6| 3189/3300 [42:17<04:43,  2.55s/it, lr: 1.0e-04 loss: 5.274e-02]
anha:  97%|#########6| 3189/3300 [42:17<04:43,  2.55s/it, lr: 1.0e-04 loss: 5.274e-02]
anha:  97%|#########6| 3189/3300 [42:18<04:43,  2.55s/it, lr: 1.0e-04 loss: 4.991e-02]
anha:  97%|#########6| 3189/3300 [42:18<04:43,  2.55s/it, lr: 1.0e-04 loss: 4.991e-02]
anha:  97%|#########6| 3190/3300 [42:20<04:40,  2.55s/it, lr: 1.0e-04 loss: 9.303e-02]
anha:  97%|#########6| 3190/3300 [42:20<04:40,  2.55s/it, lr: 1.0e-04 loss: 9.303e-02]
anha:  97%|#########6| 3191/3300 [42:20<03:19,  1.83s/it, lr: 1.0e-04 loss: 9.303e-02]
anha:  97%|#########6| 3191/3300 [42:20<03:19,  1.83s/it, lr: 1.0e-04 loss: 9.303e-02]
anha:  97%|#########6| 3191/3300 [42:23<03:19,  1.83s/it, lr: 1.0e-04 loss: 4.040e-02]
anha:  97%|#########6| 3191/3300 [42:23<03:19,  1.83s/it, lr: 1.0e-04 loss: 4.040e-02]
anha:  97%|#########6| 3192/3300 [42:23<03:59,  2.21s/it, lr: 1.0e-04 loss: 4.040e-02]
anha:  97%|#########6| 3192/3300 [42:23<03:59,  2.21s/it, lr: 1.0e-04 loss: 4.040e-02]
anha:  97%|#########6| 3192/3300 [42:26<03:59,  2.21s/it, lr: 1.0e-04 loss: 5.250e-02]
anha:  97%|#########6| 3192/3300 [42:26<03:59,  2.21s/it, lr: 1.0e-04 loss: 5.250e-02]
anha:  97%|#########6| 3193/3300 [42:26<04:31,  2.53s/it, lr: 1.0e-04 loss: 5.250e-02]
anha:  97%|#########6| 3193/3300 [42:26<04:31,  2.53s/it, lr: 1.0e-04 loss: 5.250e-02]
anha:  97%|#########6| 3193/3300 [42:30<04:31,  2.53s/it, lr: 1.0e-04 loss: 6.991e-02]
anha:  97%|#########6| 3193/3300 [42:30<04:31,  2.53s/it, lr: 1.0e-04 loss: 6.991e-02]
anha:  97%|#########6| 3194/3300 [42:30<04:52,  2.76s/it, lr: 1.0e-04 loss: 6.991e-02]
anha:  97%|#########6| 3194/3300 [42:30<04:52,  2.76s/it, lr: 1.0e-04 loss: 6.991e-02]
anha:  97%|#########6| 3194/3300 [42:32<04:52,  2.76s/it, lr: 1.0e-04 loss: 4.001e-02]
anha:  97%|#########6| 3194/3300 [42:32<04:52,  2.76s/it, lr: 1.0e-04 loss: 4.001e-02]
anha:  97%|#########6| 3195/3300 [42:32<04:19,  2.47s/it, lr: 1.0e-04 loss: 4.001e-02]
anha:  97%|#########6| 3195/3300 [42:32<04:19,  2.47s/it, lr: 1.0e-04 loss: 4.001e-02]
anha:  97%|#########6| 3195/3300 [42:34<04:19,  2.47s/it, lr: 1.0e-04 loss: 7.749e-02]
anha:  97%|#########6| 3195/3300 [42:34<04:19,  2.47s/it, lr: 1.0e-04 loss: 7.749e-02]
anha:  97%|#########6| 3196/3300 [42:34<04:03,  2.34s/it, lr: 1.0e-04 loss: 7.749e-02]
anha:  97%|#########6| 3196/3300 [42:34<04:03,  2.34s/it, lr: 1.0e-04 loss: 7.749e-02]
anha:  97%|#########6| 3196/3300 [42:37<04:03,  2.34s/it, lr: 1.0e-04 loss: 1.241e-01]
anha:  97%|#########6| 3196/3300 [42:37<04:03,  2.34s/it, lr: 1.0e-04 loss: 1.241e-01]
anha:  97%|#########6| 3197/3300 [42:37<04:32,  2.65s/it, lr: 1.0e-04 loss: 1.241e-01]
anha:  97%|#########6| 3197/3300 [42:37<04:32,  2.65s/it, lr: 1.0e-04 loss: 1.241e-01]
anha:  97%|#########6| 3197/3300 [42:39<04:32,  2.65s/it, lr: 1.0e-04 loss: 8.783e-02]
anha:  97%|#########6| 3197/3300 [42:39<04:32,  2.65s/it, lr: 1.0e-04 loss: 8.783e-02]
anha:  97%|#########6| 3198/3300 [42:39<04:10,  2.45s/it, lr: 1.0e-04 loss: 8.783e-02]
anha:  97%|#########6| 3198/3300 [42:39<04:10,  2.45s/it, lr: 1.0e-04 loss: 8.783e-02]
anha:  97%|#########6| 3198/3300 [42:41<04:10,  2.45s/it, lr: 1.0e-04 loss: 1.195e-01]
anha:  97%|#########6| 3198/3300 [42:41<04:10,  2.45s/it, lr: 1.0e-04 loss: 1.195e-01]
anha:  97%|#########6| 3199/3300 [42:41<03:53,  2.31s/it, lr: 1.0e-04 loss: 1.195e-01]
anha:  97%|#########6| 3199/3300 [42:41<03:53,  2.31s/it, lr: 1.0e-04 loss: 1.195e-01]
anha:  97%|#########6| 3199/3300 [42:42<03:53,  2.31s/it, lr: 1.0e-04 loss: 7.952e-02]
anha:  97%|#########6| 3199/3300 [42:42<03:53,  2.31s/it, lr: 1.0e-04 loss: 7.952e-02]
anha:  97%|#########6| 3200/3300 [42:44<03:51,  2.31s/it, lr: 1.0e-04 loss: 6.063e-02]
anha:  97%|#########6| 3200/3300 [42:44<03:51,  2.31s/it, lr: 1.0e-04 loss: 6.063e-02]
anha:  97%|#########7| 3201/3300 [42:44<02:48,  1.70s/it, lr: 1.0e-04 loss: 6.063e-02]
anha:  97%|#########7| 3201/3300 [42:44<02:48,  1.70s/it, lr: 1.0e-04 loss: 6.063e-02]
anha:  97%|#########7| 3201/3300 [42:46<02:48,  1.70s/it, lr: 1.0e-04 loss: 7.704e-02]
anha:  97%|#########7| 3201/3300 [42:46<02:48,  1.70s/it, lr: 1.0e-04 loss: 7.704e-02]
anha:  97%|#########7| 3202/3300 [42:46<02:47,  1.71s/it, lr: 1.0e-04 loss: 7.704e-02]
anha:  97%|#########7| 3202/3300 [42:46<02:47,  1.71s/it, lr: 1.0e-04 loss: 7.704e-02]
anha:  97%|#########7| 3202/3300 [42:47<02:47,  1.71s/it, lr: 1.0e-04 loss: 6.439e-02]
anha:  97%|#########7| 3202/3300 [42:47<02:47,  1.71s/it, lr: 1.0e-04 loss: 6.439e-02]
anha:  97%|#########7| 3203/3300 [42:47<02:46,  1.71s/it, lr: 1.0e-04 loss: 6.439e-02]
anha:  97%|#########7| 3203/3300 [42:47<02:46,  1.71s/it, lr: 1.0e-04 loss: 6.439e-02]
anha:  97%|#########7| 3203/3300 [42:49<02:46,  1.71s/it, lr: 1.0e-04 loss: 7.851e-02]
anha:  97%|#########7| 3203/3300 [42:49<02:46,  1.71s/it, lr: 1.0e-04 loss: 7.851e-02]
anha:  97%|#########7| 3204/3300 [42:49<02:51,  1.79s/it, lr: 1.0e-04 loss: 7.851e-02]
anha:  97%|#########7| 3204/3300 [42:49<02:51,  1.79s/it, lr: 1.0e-04 loss: 7.851e-02]
anha:  97%|#########7| 3204/3300 [42:53<02:51,  1.79s/it, lr: 1.0e-04 loss: 3.686e-02]
anha:  97%|#########7| 3204/3300 [42:53<02:51,  1.79s/it, lr: 1.0e-04 loss: 3.686e-02]
anha:  97%|#########7| 3205/3300 [42:53<03:31,  2.23s/it, lr: 1.0e-04 loss: 3.686e-02]
anha:  97%|#########7| 3205/3300 [42:53<03:31,  2.23s/it, lr: 1.0e-04 loss: 3.686e-02]
anha:  97%|#########7| 3205/3300 [42:56<03:31,  2.23s/it, lr: 1.0e-04 loss: 3.238e-02]
anha:  97%|#########7| 3205/3300 [42:56<03:31,  2.23s/it, lr: 1.0e-04 loss: 3.238e-02]
anha:  97%|#########7| 3206/3300 [42:56<04:01,  2.56s/it, lr: 1.0e-04 loss: 3.238e-02]
anha:  97%|#########7| 3206/3300 [42:56<04:01,  2.56s/it, lr: 1.0e-04 loss: 3.238e-02]
anha:  97%|#########7| 3206/3300 [42:57<04:01,  2.56s/it, lr: 1.0e-04 loss: 3.071e-01]
anha:  97%|#########7| 3206/3300 [42:57<04:01,  2.56s/it, lr: 1.0e-04 loss: 3.071e-01]
anha:  97%|#########7| 3207/3300 [42:57<03:16,  2.12s/it, lr: 1.0e-04 loss: 3.071e-01]
anha:  97%|#########7| 3207/3300 [42:57<03:16,  2.12s/it, lr: 1.0e-04 loss: 3.071e-01]
anha:  97%|#########7| 3207/3300 [43:00<03:16,  2.12s/it, lr: 1.0e-04 loss: 9.745e-02]
anha:  97%|#########7| 3207/3300 [43:00<03:16,  2.12s/it, lr: 1.0e-04 loss: 9.745e-02]
anha:  97%|#########7| 3208/3300 [43:00<03:25,  2.23s/it, lr: 1.0e-04 loss: 9.745e-02]
anha:  97%|#########7| 3208/3300 [43:00<03:25,  2.23s/it, lr: 1.0e-04 loss: 9.745e-02]
anha:  97%|#########7| 3208/3300 [43:01<03:25,  2.23s/it, lr: 1.0e-04 loss: 3.985e-02]
anha:  97%|#########7| 3208/3300 [43:01<03:25,  2.23s/it, lr: 1.0e-04 loss: 3.985e-02]
anha:  97%|#########7| 3209/3300 [43:01<02:50,  1.87s/it, lr: 1.0e-04 loss: 3.985e-02]
anha:  97%|#########7| 3209/3300 [43:01<02:50,  1.87s/it, lr: 1.0e-04 loss: 3.985e-02]
anha:  97%|#########7| 3209/3300 [43:02<02:50,  1.87s/it, lr: 1.0e-04 loss: 6.446e-02]
anha:  97%|#########7| 3209/3300 [43:02<02:50,  1.87s/it, lr: 1.0e-04 loss: 6.446e-02]
anha:  97%|#########7| 3210/3300 [43:03<02:48,  1.87s/it, lr: 1.0e-04 loss: 5.154e-02]
anha:  97%|#########7| 3210/3300 [43:03<02:48,  1.87s/it, lr: 1.0e-04 loss: 5.154e-02]
anha:  97%|#########7| 3211/3300 [43:03<02:05,  1.41s/it, lr: 1.0e-04 loss: 5.154e-02]
anha:  97%|#########7| 3211/3300 [43:03<02:05,  1.41s/it, lr: 1.0e-04 loss: 5.154e-02]
anha:  97%|#########7| 3211/3300 [43:05<02:05,  1.41s/it, lr: 1.0e-04 loss: 6.699e-02]
anha:  97%|#########7| 3211/3300 [43:05<02:05,  1.41s/it, lr: 1.0e-04 loss: 6.699e-02]
anha:  97%|#########7| 3212/3300 [43:05<02:10,  1.49s/it, lr: 1.0e-04 loss: 6.699e-02]
anha:  97%|#########7| 3212/3300 [43:05<02:10,  1.49s/it, lr: 1.0e-04 loss: 6.699e-02]
anha:  97%|#########7| 3212/3300 [43:06<02:10,  1.49s/it, lr: 1.0e-04 loss: 7.204e-02]
anha:  97%|#########7| 3212/3300 [43:06<02:10,  1.49s/it, lr: 1.0e-04 loss: 7.204e-02]
anha:  97%|#########7| 3213/3300 [43:06<01:58,  1.36s/it, lr: 1.0e-04 loss: 7.204e-02]
anha:  97%|#########7| 3213/3300 [43:06<01:58,  1.36s/it, lr: 1.0e-04 loss: 7.204e-02]
anha:  97%|#########7| 3213/3300 [43:10<01:58,  1.36s/it, lr: 1.0e-04 loss: 3.115e-02]
anha:  97%|#########7| 3213/3300 [43:10<01:58,  1.36s/it, lr: 1.0e-04 loss: 3.115e-02]
anha:  97%|#########7| 3214/3300 [43:10<02:46,  1.94s/it, lr: 1.0e-04 loss: 3.115e-02]
anha:  97%|#########7| 3214/3300 [43:10<02:46,  1.94s/it, lr: 1.0e-04 loss: 3.115e-02]
anha:  97%|#########7| 3214/3300 [43:12<02:46,  1.94s/it, lr: 1.0e-04 loss: 2.159e-01]
anha:  97%|#########7| 3214/3300 [43:12<02:46,  1.94s/it, lr: 1.0e-04 loss: 2.159e-01]
anha:  97%|#########7| 3215/3300 [43:12<02:47,  1.97s/it, lr: 1.0e-04 loss: 2.159e-01]
anha:  97%|#########7| 3215/3300 [43:12<02:47,  1.97s/it, lr: 1.0e-04 loss: 2.159e-01]
anha:  97%|#########7| 3215/3300 [43:13<02:47,  1.97s/it, lr: 1.0e-04 loss: 1.340e-01]
anha:  97%|#########7| 3215/3300 [43:13<02:47,  1.97s/it, lr: 1.0e-04 loss: 1.340e-01]
anha:  97%|#########7| 3216/3300 [43:13<02:22,  1.69s/it, lr: 1.0e-04 loss: 1.340e-01]
anha:  97%|#########7| 3216/3300 [43:13<02:22,  1.69s/it, lr: 1.0e-04 loss: 1.340e-01]
anha:  97%|#########7| 3216/3300 [43:14<02:22,  1.69s/it, lr: 1.0e-04 loss: 9.635e-02]
anha:  97%|#########7| 3216/3300 [43:14<02:22,  1.69s/it, lr: 1.0e-04 loss: 9.635e-02]
anha:  97%|#########7| 3217/3300 [43:14<02:04,  1.49s/it, lr: 1.0e-04 loss: 9.635e-02]
anha:  97%|#########7| 3217/3300 [43:14<02:04,  1.49s/it, lr: 1.0e-04 loss: 9.635e-02]
anha:  97%|#########7| 3217/3300 [43:15<02:04,  1.49s/it, lr: 1.0e-04 loss: 1.137e-01]
anha:  97%|#########7| 3217/3300 [43:15<02:04,  1.49s/it, lr: 1.0e-04 loss: 1.137e-01]
anha:  98%|#########7| 3218/3300 [43:15<01:50,  1.35s/it, lr: 1.0e-04 loss: 1.137e-01]
anha:  98%|#########7| 3218/3300 [43:15<01:50,  1.35s/it, lr: 1.0e-04 loss: 1.137e-01]
anha:  98%|#########7| 3218/3300 [43:16<01:50,  1.35s/it, lr: 1.0e-04 loss: 1.638e-01]
anha:  98%|#########7| 3218/3300 [43:16<01:50,  1.35s/it, lr: 1.0e-04 loss: 1.638e-01]
anha:  98%|#########7| 3219/3300 [43:16<01:40,  1.25s/it, lr: 1.0e-04 loss: 1.638e-01]
anha:  98%|#########7| 3219/3300 [43:16<01:40,  1.25s/it, lr: 1.0e-04 loss: 1.638e-01]
anha:  98%|#########7| 3219/3300 [43:17<01:40,  1.25s/it, lr: 1.0e-04 loss: 1.244e-01]
anha:  98%|#########7| 3219/3300 [43:17<01:40,  1.25s/it, lr: 1.0e-04 loss: 1.244e-01]
anha:  98%|#########7| 3220/3300 [43:18<01:39,  1.25s/it, lr: 1.0e-04 loss: 7.904e-02]
anha:  98%|#########7| 3220/3300 [43:18<01:39,  1.25s/it, lr: 1.0e-04 loss: 7.904e-02]
anha:  98%|#########7| 3221/3300 [43:18<01:11,  1.10it/s, lr: 1.0e-04 loss: 7.904e-02]
anha:  98%|#########7| 3221/3300 [43:18<01:11,  1.10it/s, lr: 1.0e-04 loss: 7.904e-02]
anha:  98%|#########7| 3221/3300 [43:19<01:11,  1.10it/s, lr: 1.0e-04 loss: 4.312e-02]
anha:  98%|#########7| 3221/3300 [43:19<01:11,  1.10it/s, lr: 1.0e-04 loss: 4.312e-02]
anha:  98%|#########7| 3222/3300 [43:19<01:12,  1.08it/s, lr: 1.0e-04 loss: 4.312e-02]
anha:  98%|#########7| 3222/3300 [43:19<01:12,  1.08it/s, lr: 1.0e-04 loss: 4.312e-02]
anha:  98%|#########7| 3222/3300 [43:20<01:12,  1.08it/s, lr: 1.0e-04 loss: 5.959e-02]
anha:  98%|#########7| 3222/3300 [43:20<01:12,  1.08it/s, lr: 1.0e-04 loss: 5.959e-02]
anha:  98%|#########7| 3223/3300 [43:20<01:12,  1.06it/s, lr: 1.0e-04 loss: 5.959e-02]
anha:  98%|#########7| 3223/3300 [43:20<01:12,  1.06it/s, lr: 1.0e-04 loss: 5.959e-02]
anha:  98%|#########7| 3223/3300 [43:22<01:12,  1.06it/s, lr: 1.0e-04 loss: 2.405e-01]
anha:  98%|#########7| 3223/3300 [43:22<01:12,  1.06it/s, lr: 1.0e-04 loss: 2.405e-01]
anha:  98%|#########7| 3224/3300 [43:22<01:33,  1.23s/it, lr: 1.0e-04 loss: 2.405e-01]
anha:  98%|#########7| 3224/3300 [43:22<01:33,  1.23s/it, lr: 1.0e-04 loss: 2.405e-01]
anha:  98%|#########7| 3224/3300 [43:23<01:33,  1.23s/it, lr: 1.0e-04 loss: 5.250e-02]
anha:  98%|#########7| 3224/3300 [43:23<01:33,  1.23s/it, lr: 1.0e-04 loss: 5.250e-02]
anha:  98%|#########7| 3225/3300 [43:23<01:30,  1.20s/it, lr: 1.0e-04 loss: 5.250e-02]
anha:  98%|#########7| 3225/3300 [43:23<01:30,  1.20s/it, lr: 1.0e-04 loss: 5.250e-02]
anha:  98%|#########7| 3225/3300 [43:26<01:30,  1.20s/it, lr: 1.0e-04 loss: 8.415e-02]
anha:  98%|#########7| 3225/3300 [43:26<01:30,  1.20s/it, lr: 1.0e-04 loss: 8.415e-02]
anha:  98%|#########7| 3226/3300 [43:26<02:14,  1.82s/it, lr: 1.0e-04 loss: 8.415e-02]
anha:  98%|#########7| 3226/3300 [43:26<02:14,  1.82s/it, lr: 1.0e-04 loss: 8.415e-02]
anha:  98%|#########7| 3226/3300 [43:28<02:14,  1.82s/it, lr: 1.0e-04 loss: 3.582e-02]
anha:  98%|#########7| 3226/3300 [43:28<02:14,  1.82s/it, lr: 1.0e-04 loss: 3.582e-02]
anha:  98%|#########7| 3227/3300 [43:28<02:15,  1.86s/it, lr: 1.0e-04 loss: 3.582e-02]
anha:  98%|#########7| 3227/3300 [43:28<02:15,  1.86s/it, lr: 1.0e-04 loss: 3.582e-02]
anha:  98%|#########7| 3227/3300 [43:30<02:15,  1.86s/it, lr: 1.0e-04 loss: 4.311e-02]
anha:  98%|#########7| 3227/3300 [43:30<02:15,  1.86s/it, lr: 1.0e-04 loss: 4.311e-02]
anha:  98%|#########7| 3228/3300 [43:30<02:15,  1.89s/it, lr: 1.0e-04 loss: 4.311e-02]
anha:  98%|#########7| 3228/3300 [43:30<02:15,  1.89s/it, lr: 1.0e-04 loss: 4.311e-02]
anha:  98%|#########7| 3228/3300 [43:31<02:15,  1.89s/it, lr: 1.0e-04 loss: 6.578e-02]
anha:  98%|#########7| 3228/3300 [43:31<02:15,  1.89s/it, lr: 1.0e-04 loss: 6.578e-02]
anha:  98%|#########7| 3229/3300 [43:31<01:55,  1.62s/it, lr: 1.0e-04 loss: 6.578e-02]
anha:  98%|#########7| 3229/3300 [43:31<01:55,  1.62s/it, lr: 1.0e-04 loss: 6.578e-02]
anha:  98%|#########7| 3229/3300 [43:32<01:55,  1.62s/it, lr: 1.0e-04 loss: 9.899e-02]
anha:  98%|#########7| 3229/3300 [43:32<01:55,  1.62s/it, lr: 1.0e-04 loss: 9.899e-02]
anha:  98%|#########7| 3230/3300 [43:34<01:53,  1.62s/it, lr: 1.0e-04 loss: 8.853e-02]
anha:  98%|#########7| 3230/3300 [43:34<01:53,  1.62s/it, lr: 1.0e-04 loss: 8.853e-02]
anha:  98%|#########7| 3231/3300 [43:34<01:28,  1.28s/it, lr: 1.0e-04 loss: 8.853e-02]
anha:  98%|#########7| 3231/3300 [43:34<01:28,  1.28s/it, lr: 1.0e-04 loss: 8.853e-02]
anha:  98%|#########7| 3231/3300 [43:37<01:28,  1.28s/it, lr: 1.0e-04 loss: 3.146e-02]
anha:  98%|#########7| 3231/3300 [43:37<01:28,  1.28s/it, lr: 1.0e-04 loss: 3.146e-02]
anha:  98%|#########7| 3232/3300 [43:37<02:01,  1.79s/it, lr: 1.0e-04 loss: 3.146e-02]
anha:  98%|#########7| 3232/3300 [43:37<02:01,  1.79s/it, lr: 1.0e-04 loss: 3.146e-02]
anha:  98%|#########7| 3232/3300 [43:38<02:01,  1.79s/it, lr: 1.0e-04 loss: 4.870e-02]
anha:  98%|#########7| 3232/3300 [43:38<02:01,  1.79s/it, lr: 1.0e-04 loss: 4.870e-02]
anha:  98%|#########7| 3233/3300 [43:38<01:46,  1.59s/it, lr: 1.0e-04 loss: 4.870e-02]
anha:  98%|#########7| 3233/3300 [43:38<01:46,  1.59s/it, lr: 1.0e-04 loss: 4.870e-02]
anha:  98%|#########7| 3233/3300 [43:40<01:46,  1.59s/it, lr: 1.0e-04 loss: 1.440e-01]
anha:  98%|#########7| 3233/3300 [43:40<01:46,  1.59s/it, lr: 1.0e-04 loss: 1.440e-01]
anha:  98%|#########8| 3234/3300 [43:40<01:54,  1.73s/it, lr: 1.0e-04 loss: 1.440e-01]
anha:  98%|#########8| 3234/3300 [43:40<01:54,  1.73s/it, lr: 1.0e-04 loss: 1.440e-01]
anha:  98%|#########8| 3234/3300 [43:44<01:54,  1.73s/it, lr: 1.0e-04 loss: 4.131e-02]
anha:  98%|#########8| 3234/3300 [43:44<01:54,  1.73s/it, lr: 1.0e-04 loss: 4.131e-02]
anha:  98%|#########8| 3235/3300 [43:44<02:24,  2.22s/it, lr: 1.0e-04 loss: 4.131e-02]
anha:  98%|#########8| 3235/3300 [43:44<02:24,  2.22s/it, lr: 1.0e-04 loss: 4.131e-02]
anha:  98%|#########8| 3235/3300 [43:47<02:24,  2.22s/it, lr: 1.0e-04 loss: 4.580e-02]
anha:  98%|#########8| 3235/3300 [43:47<02:24,  2.22s/it, lr: 1.0e-04 loss: 4.580e-02]
anha:  98%|#########8| 3236/3300 [43:47<02:42,  2.55s/it, lr: 1.0e-04 loss: 4.580e-02]
anha:  98%|#########8| 3236/3300 [43:47<02:42,  2.55s/it, lr: 1.0e-04 loss: 4.580e-02]
anha:  98%|#########8| 3236/3300 [43:48<02:42,  2.55s/it, lr: 1.0e-04 loss: 3.263e-02]
anha:  98%|#########8| 3236/3300 [43:48<02:42,  2.55s/it, lr: 1.0e-04 loss: 3.263e-02]
anha:  98%|#########8| 3237/3300 [43:48<02:12,  2.10s/it, lr: 1.0e-04 loss: 3.263e-02]
anha:  98%|#########8| 3237/3300 [43:48<02:12,  2.10s/it, lr: 1.0e-04 loss: 3.263e-02]
anha:  98%|#########8| 3237/3300 [43:49<02:12,  2.10s/it, lr: 1.0e-04 loss: 7.112e-02]
anha:  98%|#########8| 3237/3300 [43:49<02:12,  2.10s/it, lr: 1.0e-04 loss: 7.112e-02]
anha:  98%|#########8| 3238/3300 [43:49<01:50,  1.78s/it, lr: 1.0e-04 loss: 7.112e-02]
anha:  98%|#########8| 3238/3300 [43:49<01:50,  1.78s/it, lr: 1.0e-04 loss: 7.112e-02]
anha:  98%|#########8| 3238/3300 [43:51<01:50,  1.78s/it, lr: 1.0e-04 loss: 5.380e-02]
anha:  98%|#########8| 3238/3300 [43:51<01:50,  1.78s/it, lr: 1.0e-04 loss: 5.380e-02]
anha:  98%|#########8| 3239/3300 [43:51<01:47,  1.77s/it, lr: 1.0e-04 loss: 5.380e-02]
anha:  98%|#########8| 3239/3300 [43:51<01:47,  1.77s/it, lr: 1.0e-04 loss: 5.380e-02]
anha:  98%|#########8| 3239/3300 [43:53<01:47,  1.77s/it, lr: 1.0e-04 loss: 3.835e-02]
anha:  98%|#########8| 3239/3300 [43:53<01:47,  1.77s/it, lr: 1.0e-04 loss: 3.835e-02]
anha:  98%|#########8| 3240/3300 [43:54<01:45,  1.77s/it, lr: 1.0e-04 loss: 8.130e-02]
anha:  98%|#########8| 3240/3300 [43:54<01:45,  1.77s/it, lr: 1.0e-04 loss: 8.130e-02]
anha:  98%|#########8| 3241/3300 [43:54<01:09,  1.19s/it, lr: 1.0e-04 loss: 8.130e-02]
anha:  98%|#########8| 3241/3300 [43:54<01:09,  1.19s/it, lr: 1.0e-04 loss: 8.130e-02]
anha:  98%|#########8| 3241/3300 [43:56<01:09,  1.19s/it, lr: 1.0e-04 loss: 3.862e-02]
anha:  98%|#########8| 3241/3300 [43:56<01:09,  1.19s/it, lr: 1.0e-04 loss: 3.862e-02]
anha:  98%|#########8| 3242/3300 [43:56<01:16,  1.32s/it, lr: 1.0e-04 loss: 3.862e-02]
anha:  98%|#########8| 3242/3300 [43:56<01:16,  1.32s/it, lr: 1.0e-04 loss: 3.862e-02]
anha:  98%|#########8| 3242/3300 [43:57<01:16,  1.32s/it, lr: 1.0e-04 loss: 4.072e-02]
anha:  98%|#########8| 3242/3300 [43:57<01:16,  1.32s/it, lr: 1.0e-04 loss: 4.072e-02]
anha:  98%|#########8| 3243/3300 [43:57<01:10,  1.23s/it, lr: 1.0e-04 loss: 4.072e-02]
anha:  98%|#########8| 3243/3300 [43:57<01:10,  1.23s/it, lr: 1.0e-04 loss: 4.072e-02]
anha:  98%|#########8| 3243/3300 [43:59<01:10,  1.23s/it, lr: 1.0e-04 loss: 8.856e-02]
anha:  98%|#########8| 3243/3300 [43:59<01:10,  1.23s/it, lr: 1.0e-04 loss: 8.856e-02]
anha:  98%|#########8| 3244/3300 [43:59<01:21,  1.45s/it, lr: 1.0e-04 loss: 8.856e-02]
anha:  98%|#########8| 3244/3300 [43:59<01:21,  1.45s/it, lr: 1.0e-04 loss: 8.856e-02]
anha:  98%|#########8| 3244/3300 [44:02<01:21,  1.45s/it, lr: 1.0e-04 loss: 3.993e-02]
anha:  98%|#########8| 3244/3300 [44:02<01:21,  1.45s/it, lr: 1.0e-04 loss: 3.993e-02]
anha:  98%|#########8| 3245/3300 [44:02<01:50,  2.01s/it, lr: 1.0e-04 loss: 3.993e-02]
anha:  98%|#########8| 3245/3300 [44:02<01:50,  2.01s/it, lr: 1.0e-04 loss: 3.993e-02]
anha:  98%|#########8| 3245/3300 [44:04<01:50,  2.01s/it, lr: 1.0e-04 loss: 1.140e-01]
anha:  98%|#########8| 3245/3300 [44:04<01:50,  2.01s/it, lr: 1.0e-04 loss: 1.140e-01]
anha:  98%|#########8| 3246/3300 [44:04<01:48,  2.00s/it, lr: 1.0e-04 loss: 1.140e-01]
anha:  98%|#########8| 3246/3300 [44:04<01:48,  2.00s/it, lr: 1.0e-04 loss: 1.140e-01]
anha:  98%|#########8| 3246/3300 [44:07<01:48,  2.00s/it, lr: 1.0e-04 loss: 3.456e-02]
anha:  98%|#########8| 3246/3300 [44:07<01:48,  2.00s/it, lr: 1.0e-04 loss: 3.456e-02]
anha:  98%|#########8| 3247/3300 [44:07<02:05,  2.37s/it, lr: 1.0e-04 loss: 3.456e-02]
anha:  98%|#########8| 3247/3300 [44:07<02:05,  2.37s/it, lr: 1.0e-04 loss: 3.456e-02]
anha:  98%|#########8| 3247/3300 [44:11<02:05,  2.37s/it, lr: 1.0e-04 loss: 1.987e-01]
anha:  98%|#########8| 3247/3300 [44:11<02:05,  2.37s/it, lr: 1.0e-04 loss: 1.987e-01]
anha:  98%|#########8| 3248/3300 [44:11<02:18,  2.66s/it, lr: 1.0e-04 loss: 1.987e-01]
anha:  98%|#########8| 3248/3300 [44:11<02:18,  2.66s/it, lr: 1.0e-04 loss: 1.987e-01]
anha:  98%|#########8| 3248/3300 [44:12<02:18,  2.66s/it, lr: 1.0e-04 loss: 7.235e-02]
anha:  98%|#########8| 3248/3300 [44:12<02:18,  2.66s/it, lr: 1.0e-04 loss: 7.235e-02]
anha:  98%|#########8| 3249/3300 [44:12<02:01,  2.39s/it, lr: 1.0e-04 loss: 7.235e-02]
anha:  98%|#########8| 3249/3300 [44:12<02:01,  2.39s/it, lr: 1.0e-04 loss: 7.235e-02]
anha:  98%|#########8| 3249/3300 [44:13<02:01,  2.39s/it, lr: 1.0e-04 loss: 5.695e-02]
anha:  98%|#########8| 3249/3300 [44:13<02:01,  2.39s/it, lr: 1.0e-04 loss: 5.695e-02]


Generating Images: 0it [00:00, ?it/s]
Generating Images: 0it [00:00, ?it/s]


                                     
                                     
anha:  98%|#########8| 3250/3300 [44:14<01:59,  2.39s/it, lr: 1.0e-04 loss: 4.014e-02]
anha:  98%|#########8| 3250/3300 [44:14<01:59,  2.39s/it, lr: 1.0e-04 loss: 4.014e-02]
anha:  99%|#########8| 3251/3300 [44:14<01:14,  1.53s/it, lr: 1.0e-04 loss: 4.014e-02]
anha:  99%|#########8| 3251/3300 [44:14<01:14,  1.53s/it, lr: 1.0e-04 loss: 4.014e-02]
anha:  99%|#########8| 3251/3300 [44:15<01:14,  1.53s/it, lr: 1.0e-04 loss: 1.391e-01]
anha:  99%|#########8| 3251/3300 [44:15<01:14,  1.53s/it, lr: 1.0e-04 loss: 1.391e-01]
anha:  99%|#########8| 3252/3300 [44:15<01:07,  1.40s/it, lr: 1.0e-04 loss: 1.391e-01]
anha:  99%|#########8| 3252/3300 [44:15<01:07,  1.40s/it, lr: 1.0e-04 loss: 1.391e-01]
anha:  99%|#########8| 3252/3300 [44:19<01:07,  1.40s/it, lr: 1.0e-04 loss: 2.261e-01]
anha:  99%|#########8| 3252/3300 [44:19<01:07,  1.40s/it, lr: 1.0e-04 loss: 2.261e-01]
anha:  99%|#########8| 3253/3300 [44:19<01:28,  1.89s/it, lr: 1.0e-04 loss: 2.261e-01]
anha:  99%|#########8| 3253/3300 [44:19<01:28,  1.89s/it, lr: 1.0e-04 loss: 2.261e-01]
anha:  99%|#########8| 3253/3300 [44:20<01:28,  1.89s/it, lr: 1.0e-04 loss: 4.001e-02]
anha:  99%|#########8| 3253/3300 [44:20<01:28,  1.89s/it, lr: 1.0e-04 loss: 4.001e-02]
anha:  99%|#########8| 3254/3300 [44:20<01:15,  1.65s/it, lr: 1.0e-04 loss: 4.001e-02]
anha:  99%|#########8| 3254/3300 [44:20<01:15,  1.65s/it, lr: 1.0e-04 loss: 4.001e-02]
anha:  99%|#########8| 3254/3300 [44:21<01:15,  1.65s/it, lr: 1.0e-04 loss: 5.165e-02]
anha:  99%|#########8| 3254/3300 [44:21<01:15,  1.65s/it, lr: 1.0e-04 loss: 5.165e-02]
anha:  99%|#########8| 3255/3300 [44:21<01:06,  1.48s/it, lr: 1.0e-04 loss: 5.165e-02]
anha:  99%|#########8| 3255/3300 [44:21<01:06,  1.48s/it, lr: 1.0e-04 loss: 5.165e-02]
anha:  99%|#########8| 3255/3300 [44:23<01:06,  1.48s/it, lr: 1.0e-04 loss: 3.492e-02]
anha:  99%|#########8| 3255/3300 [44:23<01:06,  1.48s/it, lr: 1.0e-04 loss: 3.492e-02]
anha:  99%|#########8| 3256/3300 [44:23<01:11,  1.62s/it, lr: 1.0e-04 loss: 3.492e-02]
anha:  99%|#########8| 3256/3300 [44:23<01:11,  1.62s/it, lr: 1.0e-04 loss: 3.492e-02]
anha:  99%|#########8| 3256/3300 [44:25<01:11,  1.62s/it, lr: 1.0e-04 loss: 4.542e-02]
anha:  99%|#########8| 3256/3300 [44:25<01:11,  1.62s/it, lr: 1.0e-04 loss: 4.542e-02]
anha:  99%|#########8| 3257/3300 [44:25<01:13,  1.71s/it, lr: 1.0e-04 loss: 4.542e-02]
anha:  99%|#########8| 3257/3300 [44:25<01:13,  1.71s/it, lr: 1.0e-04 loss: 4.542e-02]
anha:  99%|#########8| 3257/3300 [44:28<01:13,  1.71s/it, lr: 1.0e-04 loss: 1.895e-01]
anha:  99%|#########8| 3257/3300 [44:28<01:13,  1.71s/it, lr: 1.0e-04 loss: 1.895e-01]
anha:  99%|#########8| 3258/3300 [44:28<01:32,  2.21s/it, lr: 1.0e-04 loss: 1.895e-01]
anha:  99%|#########8| 3258/3300 [44:28<01:32,  2.21s/it, lr: 1.0e-04 loss: 1.895e-01]
anha:  99%|#########8| 3258/3300 [44:31<01:32,  2.21s/it, lr: 1.0e-04 loss: 5.177e-02]
anha:  99%|#########8| 3258/3300 [44:31<01:32,  2.21s/it, lr: 1.0e-04 loss: 5.177e-02]
anha:  99%|#########8| 3259/3300 [44:31<01:34,  2.29s/it, lr: 1.0e-04 loss: 5.177e-02]
anha:  99%|#########8| 3259/3300 [44:31<01:34,  2.29s/it, lr: 1.0e-04 loss: 5.177e-02]
anha:  99%|#########8| 3259/3300 [44:33<01:34,  2.29s/it, lr: 1.0e-04 loss: 3.044e-02]
anha:  99%|#########8| 3259/3300 [44:33<01:34,  2.29s/it, lr: 1.0e-04 loss: 3.044e-02]
anha:  99%|#########8| 3260/3300 [44:34<01:31,  2.29s/it, lr: 1.0e-04 loss: 3.528e-02]
anha:  99%|#########8| 3260/3300 [44:34<01:31,  2.29s/it, lr: 1.0e-04 loss: 3.528e-02]
anha:  99%|#########8| 3261/3300 [44:34<00:57,  1.48s/it, lr: 1.0e-04 loss: 3.528e-02]
anha:  99%|#########8| 3261/3300 [44:34<00:57,  1.48s/it, lr: 1.0e-04 loss: 3.528e-02]
anha:  99%|#########8| 3261/3300 [44:36<00:57,  1.48s/it, lr: 1.0e-04 loss: 3.384e-02]
anha:  99%|#########8| 3261/3300 [44:36<00:57,  1.48s/it, lr: 1.0e-04 loss: 3.384e-02]
anha:  99%|#########8| 3262/3300 [44:36<01:00,  1.59s/it, lr: 1.0e-04 loss: 3.384e-02]
anha:  99%|#########8| 3262/3300 [44:36<01:00,  1.59s/it, lr: 1.0e-04 loss: 3.384e-02]
anha:  99%|#########8| 3262/3300 [44:39<01:00,  1.59s/it, lr: 1.0e-04 loss: 2.309e-01]
anha:  99%|#########8| 3262/3300 [44:39<01:00,  1.59s/it, lr: 1.0e-04 loss: 2.309e-01]
anha:  99%|#########8| 3263/3300 [44:39<01:17,  2.09s/it, lr: 1.0e-04 loss: 2.309e-01]
anha:  99%|#########8| 3263/3300 [44:39<01:17,  2.09s/it, lr: 1.0e-04 loss: 2.309e-01]
anha:  99%|#########8| 3263/3300 [44:41<01:17,  2.09s/it, lr: 1.0e-04 loss: 2.344e-02]
anha:  99%|#########8| 3263/3300 [44:41<01:17,  2.09s/it, lr: 1.0e-04 loss: 2.344e-02]
anha:  99%|#########8| 3264/3300 [44:41<01:13,  2.05s/it, lr: 1.0e-04 loss: 2.344e-02]
anha:  99%|#########8| 3264/3300 [44:41<01:13,  2.05s/it, lr: 1.0e-04 loss: 2.344e-02]
anha:  99%|#########8| 3264/3300 [44:42<01:13,  2.05s/it, lr: 1.0e-04 loss: 6.145e-02]
anha:  99%|#########8| 3264/3300 [44:42<01:13,  2.05s/it, lr: 1.0e-04 loss: 6.145e-02]
anha:  99%|#########8| 3265/3300 [44:42<01:01,  1.76s/it, lr: 1.0e-04 loss: 6.145e-02]
anha:  99%|#########8| 3265/3300 [44:42<01:01,  1.76s/it, lr: 1.0e-04 loss: 6.145e-02]
anha:  99%|#########8| 3265/3300 [44:46<01:01,  1.76s/it, lr: 1.0e-04 loss: 4.746e-02]
anha:  99%|#########8| 3265/3300 [44:46<01:01,  1.76s/it, lr: 1.0e-04 loss: 4.746e-02]
anha:  99%|#########8| 3266/3300 [44:46<01:16,  2.25s/it, lr: 1.0e-04 loss: 4.746e-02]
anha:  99%|#########8| 3266/3300 [44:46<01:16,  2.25s/it, lr: 1.0e-04 loss: 4.746e-02]
anha:  99%|#########8| 3266/3300 [44:47<01:16,  2.25s/it, lr: 1.0e-04 loss: 5.829e-02]
anha:  99%|#########8| 3266/3300 [44:47<01:16,  2.25s/it, lr: 1.0e-04 loss: 5.829e-02]
anha:  99%|#########9| 3267/3300 [44:47<01:09,  2.10s/it, lr: 1.0e-04 loss: 5.829e-02]
anha:  99%|#########9| 3267/3300 [44:47<01:09,  2.10s/it, lr: 1.0e-04 loss: 5.829e-02]
anha:  99%|#########9| 3267/3300 [44:49<01:09,  2.10s/it, lr: 1.0e-04 loss: 5.766e-02]
anha:  99%|#########9| 3267/3300 [44:49<01:09,  2.10s/it, lr: 1.0e-04 loss: 5.766e-02]
anha:  99%|#########9| 3268/3300 [44:49<01:03,  1.99s/it, lr: 1.0e-04 loss: 5.766e-02]
anha:  99%|#########9| 3268/3300 [44:49<01:03,  1.99s/it, lr: 1.0e-04 loss: 5.766e-02]
anha:  99%|#########9| 3268/3300 [44:50<01:03,  1.99s/it, lr: 1.0e-04 loss: 6.795e-02]
anha:  99%|#########9| 3268/3300 [44:50<01:03,  1.99s/it, lr: 1.0e-04 loss: 6.795e-02]
anha:  99%|#########9| 3269/3300 [44:50<00:53,  1.73s/it, lr: 1.0e-04 loss: 6.795e-02]
anha:  99%|#########9| 3269/3300 [44:50<00:53,  1.73s/it, lr: 1.0e-04 loss: 6.795e-02]
anha:  99%|#########9| 3269/3300 [44:53<00:53,  1.73s/it, lr: 1.0e-04 loss: 3.502e-02]
anha:  99%|#########9| 3269/3300 [44:53<00:53,  1.73s/it, lr: 1.0e-04 loss: 3.502e-02]
anha:  99%|#########9| 3270/3300 [44:57<00:51,  1.73s/it, lr: 1.0e-04 loss: 5.071e-02]
anha:  99%|#########9| 3270/3300 [44:57<00:51,  1.73s/it, lr: 1.0e-04 loss: 5.071e-02]
anha:  99%|#########9| 3271/3300 [44:57<00:49,  1.71s/it, lr: 1.0e-04 loss: 5.071e-02]
anha:  99%|#########9| 3271/3300 [44:57<00:49,  1.71s/it, lr: 1.0e-04 loss: 5.071e-02]
anha:  99%|#########9| 3271/3300 [44:58<00:49,  1.71s/it, lr: 1.0e-04 loss: 7.570e-02]
anha:  99%|#########9| 3271/3300 [44:58<00:49,  1.71s/it, lr: 1.0e-04 loss: 7.570e-02]
anha:  99%|#########9| 3272/3300 [44:58<00:42,  1.53s/it, lr: 1.0e-04 loss: 7.570e-02]
anha:  99%|#########9| 3272/3300 [44:58<00:42,  1.53s/it, lr: 1.0e-04 loss: 7.570e-02]
anha:  99%|#########9| 3272/3300 [45:00<00:42,  1.53s/it, lr: 1.0e-04 loss: 5.272e-02]
anha:  99%|#########9| 3272/3300 [45:00<00:42,  1.53s/it, lr: 1.0e-04 loss: 5.272e-02]
anha:  99%|#########9| 3273/3300 [45:00<00:44,  1.64s/it, lr: 1.0e-04 loss: 5.272e-02]
anha:  99%|#########9| 3273/3300 [45:00<00:44,  1.64s/it, lr: 1.0e-04 loss: 5.272e-02]
anha:  99%|#########9| 3273/3300 [45:03<00:44,  1.64s/it, lr: 1.0e-04 loss: 5.387e-01]
anha:  99%|#########9| 3273/3300 [45:03<00:44,  1.64s/it, lr: 1.0e-04 loss: 5.387e-01]
anha:  99%|#########9| 3274/3300 [45:03<00:54,  2.10s/it, lr: 1.0e-04 loss: 5.387e-01]
anha:  99%|#########9| 3274/3300 [45:03<00:54,  2.10s/it, lr: 1.0e-04 loss: 5.387e-01]
anha:  99%|#########9| 3274/3300 [45:05<00:54,  2.10s/it, lr: 1.0e-04 loss: 8.300e-02]
anha:  99%|#########9| 3274/3300 [45:05<00:54,  2.10s/it, lr: 1.0e-04 loss: 8.300e-02]
anha:  99%|#########9| 3275/3300 [45:05<00:49,  1.99s/it, lr: 1.0e-04 loss: 8.300e-02]
anha:  99%|#########9| 3275/3300 [45:05<00:49,  1.99s/it, lr: 1.0e-04 loss: 8.300e-02]
anha:  99%|#########9| 3275/3300 [45:08<00:49,  1.99s/it, lr: 1.0e-04 loss: 1.133e-01]
anha:  99%|#########9| 3275/3300 [45:08<00:49,  1.99s/it, lr: 1.0e-04 loss: 1.133e-01]
anha:  99%|#########9| 3276/3300 [45:08<00:58,  2.42s/it, lr: 1.0e-04 loss: 1.133e-01]
anha:  99%|#########9| 3276/3300 [45:08<00:58,  2.42s/it, lr: 1.0e-04 loss: 1.133e-01]
anha:  99%|#########9| 3276/3300 [45:10<00:58,  2.42s/it, lr: 1.0e-04 loss: 4.043e-02]
anha:  99%|#########9| 3276/3300 [45:10<00:58,  2.42s/it, lr: 1.0e-04 loss: 4.043e-02]
anha:  99%|#########9| 3277/3300 [45:10<00:52,  2.28s/it, lr: 1.0e-04 loss: 4.043e-02]
anha:  99%|#########9| 3277/3300 [45:10<00:52,  2.28s/it, lr: 1.0e-04 loss: 4.043e-02]
anha:  99%|#########9| 3277/3300 [45:13<00:52,  2.28s/it, lr: 1.0e-04 loss: 4.227e-02]
anha:  99%|#########9| 3277/3300 [45:13<00:52,  2.28s/it, lr: 1.0e-04 loss: 4.227e-02]
anha:  99%|#########9| 3278/3300 [45:13<00:55,  2.54s/it, lr: 1.0e-04 loss: 4.227e-02]
anha:  99%|#########9| 3278/3300 [45:13<00:55,  2.54s/it, lr: 1.0e-04 loss: 4.227e-02]
anha:  99%|#########9| 3278/3300 [45:17<00:55,  2.54s/it, lr: 1.0e-04 loss: 7.127e-02]
anha:  99%|#########9| 3278/3300 [45:17<00:55,  2.54s/it, lr: 1.0e-04 loss: 7.127e-02]
anha:  99%|#########9| 3279/3300 [45:17<00:58,  2.78s/it, lr: 1.0e-04 loss: 7.127e-02]
anha:  99%|#########9| 3279/3300 [45:17<00:58,  2.78s/it, lr: 1.0e-04 loss: 7.127e-02]
anha:  99%|#########9| 3279/3300 [45:18<00:58,  2.78s/it, lr: 1.0e-04 loss: 6.513e-02]
anha:  99%|#########9| 3279/3300 [45:18<00:58,  2.78s/it, lr: 1.0e-04 loss: 6.513e-02]
anha:  99%|#########9| 3280/3300 [45:20<00:55,  2.78s/it, lr: 1.0e-04 loss: 3.188e-02]
anha:  99%|#########9| 3280/3300 [45:20<00:55,  2.78s/it, lr: 1.0e-04 loss: 3.188e-02]
anha:  99%|#########9| 3281/3300 [45:20<00:37,  1.96s/it, lr: 1.0e-04 loss: 3.188e-02]
anha:  99%|#########9| 3281/3300 [45:20<00:37,  1.96s/it, lr: 1.0e-04 loss: 3.188e-02]
anha:  99%|#########9| 3281/3300 [45:21<00:37,  1.96s/it, lr: 1.0e-04 loss: 3.388e-02]
anha:  99%|#########9| 3281/3300 [45:21<00:37,  1.96s/it, lr: 1.0e-04 loss: 3.388e-02]
anha:  99%|#########9| 3282/3300 [45:21<00:31,  1.73s/it, lr: 1.0e-04 loss: 3.388e-02]
anha:  99%|#########9| 3282/3300 [45:21<00:31,  1.73s/it, lr: 1.0e-04 loss: 3.388e-02]
anha:  99%|#########9| 3282/3300 [45:23<00:31,  1.73s/it, lr: 1.0e-04 loss: 6.787e-02]
anha:  99%|#########9| 3282/3300 [45:23<00:31,  1.73s/it, lr: 1.0e-04 loss: 6.787e-02]
anha:  99%|#########9| 3283/3300 [45:23<00:30,  1.78s/it, lr: 1.0e-04 loss: 6.787e-02]
anha:  99%|#########9| 3283/3300 [45:23<00:30,  1.78s/it, lr: 1.0e-04 loss: 6.787e-02]
anha:  99%|#########9| 3283/3300 [45:24<00:30,  1.78s/it, lr: 1.0e-04 loss: 9.270e-02]
anha:  99%|#########9| 3283/3300 [45:24<00:30,  1.78s/it, lr: 1.0e-04 loss: 9.270e-02]
anha: 100%|#########9| 3284/3300 [45:24<00:25,  1.58s/it, lr: 1.0e-04 loss: 9.270e-02]
anha: 100%|#########9| 3284/3300 [45:24<00:25,  1.58s/it, lr: 1.0e-04 loss: 9.270e-02]
anha: 100%|#########9| 3284/3300 [45:27<00:25,  1.58s/it, lr: 1.0e-04 loss: 2.223e-01]
anha: 100%|#########9| 3284/3300 [45:27<00:25,  1.58s/it, lr: 1.0e-04 loss: 2.223e-01]
anha: 100%|#########9| 3285/3300 [45:27<00:31,  2.11s/it, lr: 1.0e-04 loss: 2.223e-01]
anha: 100%|#########9| 3285/3300 [45:27<00:31,  2.11s/it, lr: 1.0e-04 loss: 2.223e-01]
anha: 100%|#########9| 3285/3300 [45:28<00:31,  2.11s/it, lr: 1.0e-04 loss: 2.612e-01]
anha: 100%|#########9| 3285/3300 [45:28<00:31,  2.11s/it, lr: 1.0e-04 loss: 2.612e-01]
anha: 100%|#########9| 3286/3300 [45:28<00:25,  1.80s/it, lr: 1.0e-04 loss: 2.612e-01]
anha: 100%|#########9| 3286/3300 [45:28<00:25,  1.80s/it, lr: 1.0e-04 loss: 2.612e-01]
anha: 100%|#########9| 3286/3300 [45:30<00:25,  1.80s/it, lr: 1.0e-04 loss: 5.333e-02]
anha: 100%|#########9| 3286/3300 [45:30<00:25,  1.80s/it, lr: 1.0e-04 loss: 5.333e-02]
anha: 100%|#########9| 3287/3300 [45:30<00:23,  1.84s/it, lr: 1.0e-04 loss: 5.333e-02]
anha: 100%|#########9| 3287/3300 [45:30<00:23,  1.84s/it, lr: 1.0e-04 loss: 5.333e-02]
anha: 100%|#########9| 3287/3300 [45:34<00:23,  1.84s/it, lr: 1.0e-04 loss: 4.902e-01]
anha: 100%|#########9| 3287/3300 [45:34<00:23,  1.84s/it, lr: 1.0e-04 loss: 4.902e-01]
anha: 100%|#########9| 3288/3300 [45:34<00:27,  2.30s/it, lr: 1.0e-04 loss: 4.902e-01]
anha: 100%|#########9| 3288/3300 [45:34<00:27,  2.30s/it, lr: 1.0e-04 loss: 4.902e-01]
anha: 100%|#########9| 3288/3300 [45:35<00:27,  2.30s/it, lr: 1.0e-04 loss: 1.152e-01]
anha: 100%|#########9| 3288/3300 [45:35<00:27,  2.30s/it, lr: 1.0e-04 loss: 1.152e-01]
anha: 100%|#########9| 3289/3300 [45:35<00:21,  1.91s/it, lr: 1.0e-04 loss: 1.152e-01]
anha: 100%|#########9| 3289/3300 [45:35<00:21,  1.91s/it, lr: 1.0e-04 loss: 1.152e-01]
anha: 100%|#########9| 3289/3300 [45:37<00:21,  1.91s/it, lr: 1.0e-04 loss: 5.510e-02]
anha: 100%|#########9| 3289/3300 [45:37<00:21,  1.91s/it, lr: 1.0e-04 loss: 5.510e-02]
anha: 100%|#########9| 3290/3300 [45:38<00:19,  1.91s/it, lr: 1.0e-04 loss: 6.414e-02]
anha: 100%|#########9| 3290/3300 [45:38<00:19,  1.91s/it, lr: 1.0e-04 loss: 6.414e-02]
anha: 100%|#########9| 3291/3300 [45:38<00:11,  1.27s/it, lr: 1.0e-04 loss: 6.414e-02]
anha: 100%|#########9| 3291/3300 [45:38<00:11,  1.27s/it, lr: 1.0e-04 loss: 6.414e-02]
anha: 100%|#########9| 3291/3300 [45:39<00:11,  1.27s/it, lr: 1.0e-04 loss: 7.125e-02]
anha: 100%|#########9| 3291/3300 [45:39<00:11,  1.27s/it, lr: 1.0e-04 loss: 7.125e-02]
anha: 100%|#########9| 3292/3300 [45:39<00:09,  1.20s/it, lr: 1.0e-04 loss: 7.125e-02]
anha: 100%|#########9| 3292/3300 [45:39<00:09,  1.20s/it, lr: 1.0e-04 loss: 7.125e-02]
anha: 100%|#########9| 3292/3300 [45:42<00:09,  1.20s/it, lr: 1.0e-04 loss: 7.178e-02]
anha: 100%|#########9| 3292/3300 [45:42<00:09,  1.20s/it, lr: 1.0e-04 loss: 7.178e-02]
anha: 100%|#########9| 3293/3300 [45:42<00:12,  1.73s/it, lr: 1.0e-04 loss: 7.178e-02]
anha: 100%|#########9| 3293/3300 [45:42<00:12,  1.73s/it, lr: 1.0e-04 loss: 7.178e-02]
anha: 100%|#########9| 3293/3300 [45:44<00:12,  1.73s/it, lr: 1.0e-04 loss: 6.304e-02]
anha: 100%|#########9| 3293/3300 [45:44<00:12,  1.73s/it, lr: 1.0e-04 loss: 6.304e-02]
anha: 100%|#########9| 3294/3300 [45:44<00:10,  1.74s/it, lr: 1.0e-04 loss: 6.304e-02]
anha: 100%|#########9| 3294/3300 [45:44<00:10,  1.74s/it, lr: 1.0e-04 loss: 6.304e-02]
anha: 100%|#########9| 3294/3300 [45:45<00:10,  1.74s/it, lr: 1.0e-04 loss: 2.075e-01]
anha: 100%|#########9| 3294/3300 [45:45<00:10,  1.74s/it, lr: 1.0e-04 loss: 2.075e-01]
anha: 100%|#########9| 3295/3300 [45:45<00:07,  1.53s/it, lr: 1.0e-04 loss: 2.075e-01]
anha: 100%|#########9| 3295/3300 [45:45<00:07,  1.53s/it, lr: 1.0e-04 loss: 2.075e-01]
anha: 100%|#########9| 3295/3300 [45:48<00:07,  1.53s/it, lr: 1.0e-04 loss: 4.216e-02]
anha: 100%|#########9| 3295/3300 [45:48<00:07,  1.53s/it, lr: 1.0e-04 loss: 4.216e-02]
anha: 100%|#########9| 3296/3300 [45:48<00:08,  2.08s/it, lr: 1.0e-04 loss: 4.216e-02]
anha: 100%|#########9| 3296/3300 [45:48<00:08,  2.08s/it, lr: 1.0e-04 loss: 4.216e-02]
anha: 100%|#########9| 3296/3300 [45:49<00:08,  2.08s/it, lr: 1.0e-04 loss: 1.754e-01]
anha: 100%|#########9| 3296/3300 [45:49<00:08,  2.08s/it, lr: 1.0e-04 loss: 1.754e-01]
anha: 100%|#########9| 3297/3300 [45:49<00:05,  1.80s/it, lr: 1.0e-04 loss: 1.754e-01]
anha: 100%|#########9| 3297/3300 [45:49<00:05,  1.80s/it, lr: 1.0e-04 loss: 1.754e-01]
anha: 100%|#########9| 3297/3300 [45:52<00:05,  1.80s/it, lr: 1.0e-04 loss: 3.065e-02]
anha: 100%|#########9| 3297/3300 [45:52<00:05,  1.80s/it, lr: 1.0e-04 loss: 3.065e-02]
anha: 100%|#########9| 3298/3300 [45:52<00:04,  2.26s/it, lr: 1.0e-04 loss: 3.065e-02]
anha: 100%|#########9| 3298/3300 [45:52<00:04,  2.26s/it, lr: 1.0e-04 loss: 3.065e-02]
anha: 100%|#########9| 3298/3300 [45:56<00:04,  2.26s/it, lr: 1.0e-04 loss: 1.117e-01]
anha: 100%|#########9| 3298/3300 [45:56<00:04,  2.26s/it, lr: 1.0e-04 loss: 1.117e-01]
anha: 100%|#########9| 3299/3300 [45:56<00:02,  2.61s/it, lr: 1.0e-04 loss: 1.117e-01]
anha: 100%|#########9| 3299/3300 [45:56<00:02,  2.61s/it, lr: 1.0e-04 loss: 1.117e-01]
anha: 100%|#########9| 3299/3300 [45:56<00:02,  2.09s/it, lr: 1.0e-04 loss: 1.117e-01]
anha: 100%|#########9| 3299/3300 [45:56<00:02,  2.09s/it, lr: 1.0e-04 loss: 1.117e-01]


Generating Images: 0it [00:00, ?it/s]
Generating Images: 0it [00:00, ?it/s]
                                     
                                     

Saved checkpoint to /workspace/ai-toolkit/output/anha/anha.safetensors
Saved optimizer to /workspace/ai-toolkit/output/anha/optimizer.pt