Upload folder using huggingface_hub

README.md

@@ -0,0 +1,63 @@
+---
+license: apache-2.0
+base_model: t5-base
+tags:
+- summarization
+- t5
+---
+
+# HHI caption summarization model
+
+This is the summarization model from **"Learning Human-Human Interactions in Images from Weak Textual Supervision" (ICCV 2023)**: a [T5-base](https://huggingface.co/t5-base) model fine-tuned to summarize captions into short human-human interaction (HHI) descriptions. It is used to generate the pseudo-labels (pHHI) for the Who's Waldo dataset used to train the main HHI understanding model.
+
+- **Paper:** [arXiv:2304.14104](https://arxiv.org/abs/2304.14104)
+- **Code:** [github.com/tau-vailab/learning-interactions](https://github.com/tau-vailab/learning-interactions)
+- **Project page:** https://learning-interactions.github.io/
+
+## Training data
+
+Fine-tuned on synthetic caption data (`synthetic_captions.csv`, available in the [GitHub repo](https://github.com/tau-vailab/learning-interactions/blob/main/data/synthetic_captions.csv)), mapping full captions to their corresponding HHI descriptions.
+
+## Usage
+
+Can be loaded directly with `transformers`:
+
+```python
+from transformers import pipeline
+pipe = pipeline('summarization', model='malper/learning-interactions-summarization', device=0)
+pipe('summarize: ' + caption)
+```
+
+Or used with the pseudo-labeling code in the repo above (`pseudo-labeling/create_pseudolabels.py`, pass via `-m malper/learning-interactions-summarization` or after downloading locally with `hf download malper/learning-interactions-summarization --local-dir output/summarization_model`).
+
+## Training hyperparameters
+
+- learning_rate: 5e-05
+- train_batch_size: 8
+- eval_batch_size: 8
+- seed: 42
+- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
+- lr_scheduler_type: linear
+- num_epochs: 3.0
+
+### Framework versions
+
+- Transformers 4.18.0
+- PyTorch 1.13.0a0+d0d6b1f
+- Datasets 2.8.0
+- Tokenizers 0.12.1
+
+## Context
+
+This is research code from 2023, prior to the widespread availability of general-purpose vision-language models (VLMs). It is provided as-is for reproducibility of the paper's results.
+
+## Citation
+
+```bibtex
+@InProceedings{alper2023learning,
+    author    = {Morris Alper and Hadar Averbuch-Elor},
+    title     = {Learning Human-Human Interactions in Images from Weak Textual Supervision},
+    booktitle = {Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV)},
+    year      = {2023}
+}
+```

all_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 3.0,
+    "train_loss": 0.36091021529403833,
+    "train_runtime": 3139.6723,
+    "train_samples": 58895,
+    "train_samples_per_second": 56.275,
+    "train_steps_per_second": 7.034
+}

config.json

@@ -0,0 +1,58 @@
+{
+  "_name_or_path": "t5-base",
+  "architectures": [
+    "T5ForConditionalGeneration"
+  ],
+  "d_ff": 3072,
+  "d_kv": 64,
+  "d_model": 768,
+  "decoder_start_token_id": 0,
+  "dropout_rate": 0.1,
+  "eos_token_id": 1,
+  "feed_forward_proj": "relu",
+  "initializer_factor": 1.0,
+  "is_encoder_decoder": true,
+  "layer_norm_epsilon": 1e-06,
+  "model_type": "t5",
+  "n_positions": 512,
+  "num_decoder_layers": 12,
+  "num_heads": 12,
+  "num_layers": 12,
+  "output_past": true,
+  "pad_token_id": 0,
+  "relative_attention_max_distance": 128,
+  "relative_attention_num_buckets": 32,
+  "task_specific_params": {
+    "summarization": {
+      "early_stopping": true,
+      "length_penalty": 2.0,
+      "max_length": 200,
+      "min_length": 30,
+      "no_repeat_ngram_size": 3,
+      "num_beams": 4,
+      "prefix": "summarize: "
+    },
+    "translation_en_to_de": {
+      "early_stopping": true,
+      "max_length": 300,
+      "num_beams": 4,
+      "prefix": "translate English to German: "
+    },
+    "translation_en_to_fr": {
+      "early_stopping": true,
+      "max_length": 300,
+      "num_beams": 4,
+      "prefix": "translate English to French: "
+    },
+    "translation_en_to_ro": {
+      "early_stopping": true,
+      "max_length": 300,
+      "num_beams": 4,
+      "prefix": "translate English to Romanian: "
+    }
+  },
+  "torch_dtype": "float32",
+  "transformers_version": "4.18.0",
+  "use_cache": true,
+  "vocab_size": 32100
+}

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:196bddb03c00374fdb885ce919c511f7e8491e7496c7c75047a9d28143072575
+size 891616913

special_tokens_map.json

@@ -0,0 +1 @@
+{"eos_token": "</s>", "unk_token": "<unk>", "pad_token": "<pad>", "additional_special_tokens": ["<extra_id_0>", "<extra_id_1>", "<extra_id_2>", "<extra_id_3>", "<extra_id_4>", "<extra_id_5>", "<extra_id_6>", "<extra_id_7>", "<extra_id_8>", "<extra_id_9>", "<extra_id_10>", "<extra_id_11>", "<extra_id_12>", "<extra_id_13>", "<extra_id_14>", "<extra_id_15>", "<extra_id_16>", "<extra_id_17>", "<extra_id_18>", "<extra_id_19>", "<extra_id_20>", "<extra_id_21>", "<extra_id_22>", "<extra_id_23>", "<extra_id_24>", "<extra_id_25>", "<extra_id_26>", "<extra_id_27>", "<extra_id_28>", "<extra_id_29>", "<extra_id_30>", "<extra_id_31>", "<extra_id_32>", "<extra_id_33>", "<extra_id_34>", "<extra_id_35>", "<extra_id_36>", "<extra_id_37>", "<extra_id_38>", "<extra_id_39>", "<extra_id_40>", "<extra_id_41>", "<extra_id_42>", "<extra_id_43>", "<extra_id_44>", "<extra_id_45>", "<extra_id_46>", "<extra_id_47>", "<extra_id_48>", "<extra_id_49>", "<extra_id_50>", "<extra_id_51>", "<extra_id_52>", "<extra_id_53>", "<extra_id_54>", "<extra_id_55>", "<extra_id_56>", "<extra_id_57>", "<extra_id_58>", "<extra_id_59>", "<extra_id_60>", "<extra_id_61>", "<extra_id_62>", "<extra_id_63>", "<extra_id_64>", "<extra_id_65>", "<extra_id_66>", "<extra_id_67>", "<extra_id_68>", "<extra_id_69>", "<extra_id_70>", "<extra_id_71>", "<extra_id_72>", "<extra_id_73>", "<extra_id_74>", "<extra_id_75>", "<extra_id_76>", "<extra_id_77>", "<extra_id_78>", "<extra_id_79>", "<extra_id_80>", "<extra_id_81>", "<extra_id_82>", "<extra_id_83>", "<extra_id_84>", "<extra_id_85>", "<extra_id_86>", "<extra_id_87>", "<extra_id_88>", "<extra_id_89>", "<extra_id_90>", "<extra_id_91>", "<extra_id_92>", "<extra_id_93>", "<extra_id_94>", "<extra_id_95>", "<extra_id_96>", "<extra_id_97>", "<extra_id_98>", "<extra_id_99>"]}

spiece.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d60acb128cf7b7f2536e8f38a5b18a05535c9e14c7a355904270e15b0945ea86
+size 791656

tokenizer_config.json

@@ -0,0 +1 @@
+{"eos_token": "</s>", "unk_token": "<unk>", "pad_token": "<pad>", "extra_ids": 100, "additional_special_tokens": ["<extra_id_0>", "<extra_id_1>", "<extra_id_2>", "<extra_id_3>", "<extra_id_4>", "<extra_id_5>", "<extra_id_6>", "<extra_id_7>", "<extra_id_8>", "<extra_id_9>", "<extra_id_10>", "<extra_id_11>", "<extra_id_12>", "<extra_id_13>", "<extra_id_14>", "<extra_id_15>", "<extra_id_16>", "<extra_id_17>", "<extra_id_18>", "<extra_id_19>", "<extra_id_20>", "<extra_id_21>", "<extra_id_22>", "<extra_id_23>", "<extra_id_24>", "<extra_id_25>", "<extra_id_26>", "<extra_id_27>", "<extra_id_28>", "<extra_id_29>", "<extra_id_30>", "<extra_id_31>", "<extra_id_32>", "<extra_id_33>", "<extra_id_34>", "<extra_id_35>", "<extra_id_36>", "<extra_id_37>", "<extra_id_38>", "<extra_id_39>", "<extra_id_40>", "<extra_id_41>", "<extra_id_42>", "<extra_id_43>", "<extra_id_44>", "<extra_id_45>", "<extra_id_46>", "<extra_id_47>", "<extra_id_48>", "<extra_id_49>", "<extra_id_50>", "<extra_id_51>", "<extra_id_52>", "<extra_id_53>", "<extra_id_54>", "<extra_id_55>", "<extra_id_56>", "<extra_id_57>", "<extra_id_58>", "<extra_id_59>", "<extra_id_60>", "<extra_id_61>", "<extra_id_62>", "<extra_id_63>", "<extra_id_64>", "<extra_id_65>", "<extra_id_66>", "<extra_id_67>", "<extra_id_68>", "<extra_id_69>", "<extra_id_70>", "<extra_id_71>", "<extra_id_72>", "<extra_id_73>", "<extra_id_74>", "<extra_id_75>", "<extra_id_76>", "<extra_id_77>", "<extra_id_78>", "<extra_id_79>", "<extra_id_80>", "<extra_id_81>", "<extra_id_82>", "<extra_id_83>", "<extra_id_84>", "<extra_id_85>", "<extra_id_86>", "<extra_id_87>", "<extra_id_88>", "<extra_id_89>", "<extra_id_90>", "<extra_id_91>", "<extra_id_92>", "<extra_id_93>", "<extra_id_94>", "<extra_id_95>", "<extra_id_96>", "<extra_id_97>", "<extra_id_98>", "<extra_id_99>"], "model_max_length": 512, "special_tokens_map_file": null, "name_or_path": "t5-base", "tokenizer_class": "T5Tokenizer"}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 3.0,
+    "train_loss": 0.36091021529403833,
+    "train_runtime": 3139.6723,
+    "train_samples": 58895,
+    "train_samples_per_second": 56.275,
+    "train_steps_per_second": 7.034
+}

trainer_state.json

@@ -0,0 +1,289 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "global_step": 22086,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.8868061215249484e-05,
+      "loss": 0.6807,
+      "step": 500
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.773612243049896e-05,
+      "loss": 0.4855,
+      "step": 1000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.660418364574844e-05,
+      "loss": 0.4476,
+      "step": 1500
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.547224486099792e-05,
+      "loss": 0.4386,
+      "step": 2000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.43403060762474e-05,
+      "loss": 0.4359,
+      "step": 2500
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.3208367291496876e-05,
+      "loss": 0.4146,
+      "step": 3000
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.207642850674636e-05,
+      "loss": 0.4143,
+      "step": 3500
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.094448972199584e-05,
+      "loss": 0.3972,
+      "step": 4000
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 3.981255093724531e-05,
+      "loss": 0.4009,
+      "step": 4500
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 3.8680612152494794e-05,
+      "loss": 0.4085,
+      "step": 5000
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.7548673367744275e-05,
+      "loss": 0.3819,
+      "step": 5500
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.6416734582993756e-05,
+      "loss": 0.394,
+      "step": 6000
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.528479579824323e-05,
+      "loss": 0.3901,
+      "step": 6500
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.415285701349271e-05,
+      "loss": 0.3975,
+      "step": 7000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.302091822874219e-05,
+      "loss": 0.3761,
+      "step": 7500
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 3.188897944399167e-05,
+      "loss": 0.3519,
+      "step": 8000
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 3.075704065924115e-05,
+      "loss": 0.3385,
+      "step": 8500
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.962510187449063e-05,
+      "loss": 0.3505,
+      "step": 9000
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 2.849316308974011e-05,
+      "loss": 0.3364,
+      "step": 9500
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 2.7361224304989585e-05,
+      "loss": 0.347,
+      "step": 10000
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 2.6229285520239066e-05,
+      "loss": 0.3378,
+      "step": 10500
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 2.5097346735488547e-05,
+      "loss": 0.3523,
+      "step": 11000
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.3965407950738025e-05,
+      "loss": 0.3441,
+      "step": 11500
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.2833469165987503e-05,
+      "loss": 0.3493,
+      "step": 12000
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2.1701530381236984e-05,
+      "loss": 0.3416,
+      "step": 12500
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 2.0569591596486462e-05,
+      "loss": 0.3364,
+      "step": 13000
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 1.9437652811735943e-05,
+      "loss": 0.3401,
+      "step": 13500
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.830571402698542e-05,
+      "loss": 0.3543,
+      "step": 14000
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.7173775242234902e-05,
+      "loss": 0.3269,
+      "step": 14500
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 1.604183645748438e-05,
+      "loss": 0.3228,
+      "step": 15000
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 1.4909897672733858e-05,
+      "loss": 0.3102,
+      "step": 15500
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 1.3777958887983339e-05,
+      "loss": 0.3203,
+      "step": 16000
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 1.2646020103232817e-05,
+      "loss": 0.2996,
+      "step": 16500
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 1.1514081318482298e-05,
+      "loss": 0.3019,
+      "step": 17000
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 1.0382142533731776e-05,
+      "loss": 0.3116,
+      "step": 17500
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 9.250203748981255e-06,
+      "loss": 0.2996,
+      "step": 18000
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 8.118264964230735e-06,
+      "loss": 0.3065,
+      "step": 18500
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 6.986326179480214e-06,
+      "loss": 0.3168,
+      "step": 19000
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 5.8543873947296935e-06,
+      "loss": 0.3124,
+      "step": 19500
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 4.722448609979173e-06,
+      "loss": 0.2943,
+      "step": 20000
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 3.5905098252286516e-06,
+      "loss": 0.3034,
+      "step": 20500
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 2.458571040478131e-06,
+      "loss": 0.2954,
+      "step": 21000
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 1.3266322557276103e-06,
+      "loss": 0.3102,
+      "step": 21500
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 1.9469347097708958e-07,
+      "loss": 0.3105,
+      "step": 22000
+    },
+    {
+      "epoch": 3.0,
+      "step": 22086,
+      "total_flos": 1.617792737439744e+16,
+      "train_loss": 0.36091021529403833,
+      "train_runtime": 3139.6723,
+      "train_samples_per_second": 56.275,
+      "train_steps_per_second": 7.034
+    }
+  ],
+  "max_steps": 22086,
+  "num_train_epochs": 3,
+  "total_flos": 1.617792737439744e+16,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2893115af2a83bf104b73befa037acbcc1f97bd339deae8d2ac3cb3f0a8c6faf
+size 3195