evalscope 0.14.0__py3-none-any.whl → 0.15.1__py3-none-any.whl

evalscope/metrics/t2v_metrics/models/vqascore_models/lavis/configs/models/blip2/blip2_pretrain.yaml

@@ -0,0 +1,36 @@
+ # Copyright (c) 2022, salesforce.com, inc.
+ # All rights reserved.
+ # SPDX-License-Identifier: BSD-3-Clause
+ # For full license text, see the LICENSE file in the repo root or https://opensource.org/licenses/BSD-3-Clause
+
+model:
+  arch: pretrain
+  load_finetuned: False
+
+  pretrained: "https://storage.googleapis.com/sfr-vision-language-research/LAVIS/models/BLIP2/blip2_pretrained.pth"
+  finetuned: ""
+
+  # vit encoder
+  image_size: 224
+  drop_path_rate: 0
+  use_grad_checkpoint: False
+  vit_precision: "fp16"
+  freeze_vit: True
+
+  # Q-Former
+  num_query_token: 32
+
+
+preprocess:
+    vis_processor:
+        train:
+          name: "blip_image_train"
+          image_size: 224
+        eval:
+          name: "blip_image_eval"
+          image_size: 224
+    text_processor:
+        train:
+          name: "blip_caption"
+        eval:
+          name: "blip_caption"

evalscope/metrics/t2v_metrics/models/vqascore_models/lavis/configs/models/blip2/blip2_pretrain_flant5xl.yaml

@@ -0,0 +1,42 @@
+ # Copyright (c) 2022, salesforce.com, inc.
+ # All rights reserved.
+ # SPDX-License-Identifier: BSD-3-Clause
+ # For full license text, see the LICENSE file in the repo root or https://opensource.org/licenses/BSD-3-Clause
+
+model:
+  arch: pretrain_flant5xl
+  load_finetuned: False
+
+  pretrained: "https://storage.googleapis.com/sfr-vision-language-research/LAVIS/models/BLIP2/blip2_pretrained_flant5xl.pth"
+  finetuned: ""
+
+  # vit encoder
+  image_size: 224
+  drop_path_rate: 0
+  use_grad_checkpoint: False
+  vit_precision: "fp16"
+  freeze_vit: True
+
+  # Q-Former
+  num_query_token: 32
+
+  # T5
+  t5_model: "google/flan-t5-xl"
+
+  # generation configs
+  prompt: ""
+
+
+preprocess:
+    vis_processor:
+        train:
+          name: "blip_image_train"
+          image_size: 224
+        eval:
+          name: "blip_image_eval"
+          image_size: 224
+    text_processor:
+        train:
+          name: "blip_caption"
+        eval:
+          name: "blip_caption"

evalscope/metrics/t2v_metrics/models/vqascore_models/lavis/configs/models/blip2/blip2_pretrain_flant5xl_iter_80k_total_100k_no_prefix.yaml

@@ -0,0 +1,42 @@
+ # Copyright (c) 2022, salesforce.com, inc.
+ # All rights reserved.
+ # SPDX-License-Identifier: BSD-3-Clause
+ # For full license text, see the LICENSE file in the repo root or https://opensource.org/licenses/BSD-3-Clause
+
+model:
+  arch: pretrain_flant5xl
+  load_finetuned: False
+
+  pretrained: "lavis/output/BLIP2/Pretrain_stage2_flant5_xl_batch_80_no_prefix_iter_100000/20231015004/checkpoint_80000.pth"
+  finetuned: ""
+
+  # vit encoder
+  image_size: 224
+  drop_path_rate: 0
+  use_grad_checkpoint: False
+  vit_precision: "fp16"
+  freeze_vit: True
+
+  # Q-Former
+  num_query_token: 32
+
+  # T5
+  t5_model: "google/flan-t5-xl"
+
+  # generation configs
+  prompt: ""
+
+
+preprocess:
+    vis_processor:
+        train:
+          name: "blip_image_train"
+          image_size: 224
+        eval:
+          name: "blip_image_eval"
+          image_size: 224
+    text_processor:
+        train:
+          name: "blip_caption"
+        eval:
+          name: "blip_caption"

evalscope/metrics/t2v_metrics/models/vqascore_models/lavis/configs/models/blip2/blip2_pretrain_flant5xl_iter_80k_total_100k_prefix.yaml

@@ -0,0 +1,42 @@
+ # Copyright (c) 2022, salesforce.com, inc.
+ # All rights reserved.
+ # SPDX-License-Identifier: BSD-3-Clause
+ # For full license text, see the LICENSE file in the repo root or https://opensource.org/licenses/BSD-3-Clause
+
+model:
+  arch: pretrain_flant5xl
+  load_finetuned: False
+
+  pretrained: "lavis/output/BLIP2/Pretrain_stage2_flant5_xl_batch_80_prefix_iter_100000/20231015004/checkpoint_80000.pth"
+  finetuned: ""
+
+  # vit encoder
+  image_size: 224
+  drop_path_rate: 0
+  use_grad_checkpoint: False
+  vit_precision: "fp16"
+  freeze_vit: True
+
+  # Q-Former
+  num_query_token: 32
+
+  # T5
+  t5_model: "google/flan-t5-xl"
+
+  # generation configs
+  prompt: ""
+
+
+preprocess:
+    vis_processor:
+        train:
+          name: "blip_image_train"
+          image_size: 224
+        eval:
+          name: "blip_image_eval"
+          image_size: 224
+    text_processor:
+        train:
+          name: "blip_caption"
+        eval:
+          name: "blip_caption"

evalscope/metrics/t2v_metrics/models/vqascore_models/lavis/configs/models/blip2/blip2_pretrain_flant5xl_vitL.yaml

@@ -0,0 +1,43 @@
+ # Copyright (c) 2022, salesforce.com, inc.
+ # All rights reserved.
+ # SPDX-License-Identifier: BSD-3-Clause
+ # For full license text, see the LICENSE file in the repo root or https://opensource.org/licenses/BSD-3-Clause
+
+model:
+  arch: pretrain_flant5xl
+  load_finetuned: False
+
+  pretrained: "https://storage.googleapis.com/sfr-vision-language-research/LAVIS/models/BLIP2/blip2_pretrained_flant5xl_vitL.pth"
+  finetuned: ""
+
+  # vit encoder
+  vit_model: "clip_L"
+  image_size: 224
+  drop_path_rate: 0
+  use_grad_checkpoint: False
+  vit_precision: "fp16"
+  freeze_vit: True
+
+  # Q-Former
+  num_query_token: 32
+
+  # T5
+  t5_model: "google/flan-t5-xl"
+
+  # generation configs
+  prompt: ""
+
+
+preprocess:
+    vis_processor:
+        train:
+          name: "blip_image_train"
+          image_size: 224
+        eval:
+          name: "blip_image_eval"
+          image_size: 224
+    text_processor:
+        train:
+          name: "blip_caption"
+        eval:
+          name: "blip_caption"

evalscope/metrics/t2v_metrics/models/vqascore_models/lavis/configs/models/blip2/blip2_pretrain_flant5xxl.yaml

@@ -0,0 +1,42 @@
+ # Copyright (c) 2022, salesforce.com, inc.
+ # All rights reserved.
+ # SPDX-License-Identifier: BSD-3-Clause
+ # For full license text, see the LICENSE file in the repo root or https://opensource.org/licenses/BSD-3-Clause
+
+model:
+  arch: pretrain_flant5xxl
+  load_finetuned: False
+
+  pretrained: "https://storage.googleapis.com/sfr-vision-language-research/LAVIS/models/BLIP2/blip2_pretrained_flant5xxl.pth"
+  finetuned: ""
+
+  # vit encoder
+  image_size: 224
+  drop_path_rate: 0
+  use_grad_checkpoint: False
+  vit_precision: "fp16"
+  freeze_vit: True
+
+  # Q-Former
+  num_query_token: 32
+
+  # T5
+  t5_model: "google/flan-t5-xxl"
+
+  # generation configs
+  prompt: ""
+
+
+preprocess:
+    vis_processor:
+        train:
+          name: "blip_image_train"
+          image_size: 224
+        eval:
+          name: "blip_image_eval"
+          image_size: 224
+    text_processor:
+        train:
+          name: "blip_caption"
+        eval:
+          name: "blip_caption"

evalscope/metrics/t2v_metrics/models/vqascore_models/lavis/configs/models/blip2/blip2_pretrain_opt2.7b.yaml

@@ -0,0 +1,42 @@
+ # Copyright (c) 2022, salesforce.com, inc.
+ # All rights reserved.
+ # SPDX-License-Identifier: BSD-3-Clause
+ # For full license text, see the LICENSE file in the repo root or https://opensource.org/licenses/BSD-3-Clause
+
+model:
+  arch: pretrain_opt2.7b
+  load_finetuned: False
+
+  pretrained: "https://storage.googleapis.com/sfr-vision-language-research/LAVIS/models/BLIP2/blip2_pretrained_opt2.7b.pth"
+  finetuned: ""
+
+  # vit encoder
+  image_size: 224
+  drop_path_rate: 0
+  use_grad_checkpoint: False
+  vit_precision: "fp16"
+  freeze_vit: True
+
+  # Q-Former
+  num_query_token: 32
+
+  # OPT
+  opt_model: "facebook/opt-2.7b"
+
+  # generation configs
+  prompt: ""
+
+
+preprocess:
+    vis_processor:
+        train:
+          name: "blip_image_train"
+          image_size: 224
+        eval:
+          name: "blip_image_eval"
+          image_size: 224
+    text_processor:
+        train:
+          name: "blip_caption"
+        eval:
+          name: "blip_caption"

evalscope/metrics/t2v_metrics/models/vqascore_models/lavis/configs/models/blip2/blip2_pretrain_opt6.7b.yaml

@@ -0,0 +1,42 @@
+ # Copyright (c) 2022, salesforce.com, inc.
+ # All rights reserved.
+ # SPDX-License-Identifier: BSD-3-Clause
+ # For full license text, see the LICENSE file in the repo root or https://opensource.org/licenses/BSD-3-Clause
+
+model:
+  arch: pretrain_opt6.7b
+  load_finetuned: False
+
+  pretrained: "https://storage.googleapis.com/sfr-vision-language-research/LAVIS/models/BLIP2/blip2_pretrained_opt6.7b.pth"
+  finetuned: ""
+
+  # vit encoder
+  image_size: 224
+  drop_path_rate: 0
+  use_grad_checkpoint: False
+  vit_precision: "fp16"
+  freeze_vit: True
+
+  # Q-Former
+  num_query_token: 32
+
+  # OPT
+  opt_model: "facebook/opt-6.7b"
+
+  # generation configs
+  prompt: ""
+
+
+preprocess:
+    vis_processor:
+        train:
+          name: "blip_image_train"
+          image_size: 224
+        eval:
+          name: "blip_image_eval"
+          image_size: 224
+    text_processor:
+        train:
+          name: "blip_caption"
+        eval:
+          name: "blip_caption"

evalscope/metrics/t2v_metrics/models/vqascore_models/lavis/configs/models/blip2/blip2_pretrain_vitL.yaml

@@ -0,0 +1,37 @@
+ # Copyright (c) 2022, salesforce.com, inc.
+ # All rights reserved.
+ # SPDX-License-Identifier: BSD-3-Clause
+ # For full license text, see the LICENSE file in the repo root or https://opensource.org/licenses/BSD-3-Clause
+
+model:
+  arch: pretrain
+  load_finetuned: False
+
+  pretrained: "https://storage.googleapis.com/sfr-vision-language-research/LAVIS/models/BLIP2/blip2_pretrained_vitL.pth"
+  finetuned: ""
+
+  # vit encoder
+  vit_model: "clip_L"
+  image_size: 224
+  drop_path_rate: 0
+  use_grad_checkpoint: False
+  vit_precision: "fp16"
+  freeze_vit: True
+
+  # Q-Former
+  num_query_token: 32
+
+
+preprocess:
+    vis_processor:
+        train:
+          name: "blip_image_train"
+          image_size: 224
+        eval:
+          name: "blip_image_eval"
+          image_size: 224
+    text_processor:
+        train:
+          name: "blip_caption"
+        eval:
+          name: "blip_caption"

evalscope/metrics/t2v_metrics/models/vqascore_models/lavis/configs/models/blip2/blip2_vicuna13b.yaml

@@ -0,0 +1,43 @@
+# Copyright (c) 2022, salesforce.com, inc.
+# All rights reserved.
+# SPDX-License-Identifier: BSD-3-Clause
+# For full license text, see the LICENSE file in the repo root or https://opensource.org/licenses/BSD-3-Clause
+
+model:
+ arch: blip2_vicuna13b
+ load_finetuned: False
+ load_pretrained: True
+
+ pretrained: "https://storage.googleapis.com/sfr-vision-language-research/LAVIS/models/BLIP2/blip2_pretrained_vicuna13b.pth"
+ finetuned: ""
+
+ # vit encoder
+ image_size: 224
+ drop_path_rate: 0
+ use_grad_checkpoint: False
+ vit_precision: "fp16"
+ freeze_vit: True
+
+ # Q-Former
+ num_query_token: 32
+
+ # path to Vicuna checkpoint
+ llm_model: "lmsys/vicuna-13b-v1.1"
+
+ # generation configs
+ prompt: ""
+
+
+preprocess:
+   vis_processor:
+       train:
+         name: "blip2_image_train"
+         image_size: 224
+       eval:
+         name: "blip_image_eval"
+         image_size: 224
+   text_processor:
+       train:
+         name: "blip_caption"
+       eval:
+         name: "blip_caption"

evalscope/metrics/t2v_metrics/models/vqascore_models/lavis/configs/models/blip2/blip2_vicuna7b.yaml

@@ -0,0 +1,43 @@
+ # Copyright (c) 2022, salesforce.com, inc.
+ # All rights reserved.
+ # SPDX-License-Identifier: BSD-3-Clause
+ # For full license text, see the LICENSE file in the repo root or https://opensource.org/licenses/BSD-3-Clause
+
+model:
+  arch: blip2_vicuna7b
+  load_finetuned: False
+  load_pretrained: True
+
+  pretrained: "https://storage.googleapis.com/sfr-vision-language-research/LAVIS/models/BLIP2/blip2_pretrained_vicuna7b.pth"
+  finetuned: ""
+
+  # vit encoder
+  image_size: 224
+  drop_path_rate: 0
+  use_grad_checkpoint: False
+  vit_precision: "fp16"
+  freeze_vit: True
+
+  # Q-Former
+  num_query_token: 32
+
+  # path to Vicuna checkpoint
+  llm_model: "lmsys/vicuna-7b-v1.1"
+
+  # generation configs
+  prompt: ""
+
+
+preprocess:
+    vis_processor:
+        train:
+          name: "blip2_image_train"
+          image_size: 224
+        eval:
+          name: "blip_image_eval"
+          image_size: 224
+    text_processor:
+        train:
+          name: "blip_caption"
+        eval:
+          name: "blip_caption"

evalscope/metrics/t2v_metrics/models/vqascore_models/lavis/configs/models/med_config.json

@@ -0,0 +1,21 @@
+{
+  "architectures": [
+    "BertModel"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 768,
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "layer_norm_eps": 1e-12,
+  "max_position_embeddings": 512,
+  "model_type": "bert",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 12,
+  "pad_token_id": 0,
+  "add_type_embeddings": false,
+  "vocab_size": 30524,
+  "encoder_width": 768,
+  "add_cross_attention": true
+}

evalscope/metrics/t2v_metrics/models/vqascore_models/lavis/configs/models/med_config_albef.json

@@ -0,0 +1,22 @@
+{
+  "architectures": [
+    "BertModel"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 768,
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "layer_norm_eps": 1e-12,
+  "max_position_embeddings": 512,
+  "model_type": "bert",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 12,
+  "pad_token_id": 0,
+  "add_type_embeddings": false,
+  "vocab_size": 30522,
+  "encoder_width": 768,
+  "add_cross_attention": true,
+  "fusion_layer": 6
+}

evalscope/metrics/t2v_metrics/models/vqascore_models/lavis/configs/models/med_large_config.json

@@ -0,0 +1,21 @@
+{
+  "architectures": [
+    "BertModel"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 768,
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "layer_norm_eps": 1e-12,
+  "max_position_embeddings": 512,
+  "model_type": "bert",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 12,
+  "pad_token_id": 0,
+  "add_type_embeddings": false,
+  "vocab_size": 30524,
+  "encoder_width": 1024,
+  "add_cross_attention": true
+}