This is a Cross Encoder model finetuned from Qwen/Qwen3.5-0.8B on the image_to_text and text_to_image datasets using the sentence-transformers library. It computes scores for pairs of texts, which can be used for text reranking and semantic search.

Model Details

Model Description

Model Type: Cross Encoder
Base model: Qwen/Qwen3.5-0.8B
Maximum Sequence Length: 262144 tokens
Number of Output Labels: 1 label
Supported Modalities: Text, Image, Video, Message
Training Datasets:
- image_to_text
- text_to_image

Model Sources

Documentation: Sentence Transformers Documentation
Documentation: Cross Encoder Documentation
Repository: Sentence Transformers on GitHub
Hugging Face: Cross Encoders on Hugging Face

Full Model Architecture

CrossEncoder(
  (0): Transformer({'transformer_task': 'any-to-any', 'modality_config': {'text': {'method': 'forward', 'method_output_name': 'logits'}, 'image': {'method': 'forward', 'method_output_name': 'logits'}, 'video': {'method': 'forward', 'method_output_name': 'logits'}, 'message': {'method': 'forward', 'method_output_name': 'logits', 'format': 'structured'}}, 'module_output_name': 'causal_logits', 'processing_kwargs': {'chat_template': {'add_generation_prompt': True}}, 'architecture': 'Qwen3_5ForConditionalGeneration'})
  (1): LogitScore({'true_token_id': 16, 'false_token_id': 15, 'module_input_name': 'causal_logits'})
)

Usage

Direct Usage (Sentence Transformers)

First install the Sentence Transformers library:

pip install -U sentence-transformers

Then you can load this model and run inference.

python

from sentence_transformers import CrossEncoder

# Download from the 🤗 Hub
model = CrossEncoder("kwondw/reranker-Qwen3.5-0.8B-doodles-any-to-any")
# Get scores for pairs of inputs
pairs = [
    ['https://huggingface.co/kwondw/reranker-Qwen3.5-0.8B-doodles-any-to-any/resolve/main/assets/image_0.jpg', 'a content character with a tan head and purple puffballs hair wearing a blue fleece, green background'],
    ['https://huggingface.co/kwondw/reranker-Qwen3.5-0.8B-doodles-any-to-any/resolve/main/assets/image_0.jpg', 'a mustache character with a orange head and yellow headband hair wearing a white turtleneck, gradient 3 background'],
    ['https://huggingface.co/kwondw/reranker-Qwen3.5-0.8B-doodles-any-to-any/resolve/main/assets/image_0.jpg', 'a sunglasses character with a purple head and yellow bowlcut hair wearing a orange collar, gradient 4 background'],
    ['https://huggingface.co/kwondw/reranker-Qwen3.5-0.8B-doodles-any-to-any/resolve/main/assets/image_0.jpg', 'a mad note character with a orange head and pink toque hair wearing a orange puffer, grey background'],
    ['https://huggingface.co/kwondw/reranker-Qwen3.5-0.8B-doodles-any-to-any/resolve/main/assets/image_0.jpg', 'a pink beard character with a yellow head and green bowlcut hair wearing a striped sweater, light blue background'],
]
scores = model.predict(pairs)
print(scores)
# [0.8176 0.6298 0.6689 0.7432 0.7356]

Evaluation

Metrics

Cross Encoder Reranking

Datasets: doodles-image-to-text-eval and doodles-text-to-image-eval
Evaluated with CrossEncoderRerankingEvaluator with these parameters:
json
1{ 2 "at_k": 10 3}

Metric	doodles-image-to-text-eval	doodles-text-to-image-eval
map	0.9825	0.755
mrr@10	0.9825	0.755
ndcg@10	0.9869	0.8167

Training Details

Training Datasets

image_to_text

Dataset: image_to_text at a575ac6
Size: 4,500 training samples
Columns: image, text, and label
Approximate statistics based on the first 100 samples:
image text label
type image string int
modality image text
details
min: 128x128 px
mean: 128x128 px
max: 128x128 px
min: 30 tokens
mean: 33.48 tokens
max: 37 tokens
0: ~79.81%
1: ~20.19%

Samples:

image	text	label
	`a cobain glasses character with a gradient 2 head and purple puffballs hair wearing a white sweater, gradient 4 background`	`1`
	`a content character with a orange head and purple long hair wearing a striped sweater, yellow background`	`0`
	`a neutral note character with a orange head and green puffballs hair wearing a combo 2 puffer, light blue background`	`0`

Loss: BinaryCrossEntropyLoss with these parameters:
json
1{ 2 "activation_fn": "torch.nn.modules.linear.Identity", 3 "pos_weight": null 4}

text_to_image

Dataset: text_to_image at a575ac6
Size: 4,500 training samples
Columns: text, image, and label
Approximate statistics based on the first 100 samples:
text image label
type string image int
modality text image
details
min: 30 tokens
mean: 33.13 tokens
max: 37 tokens
min: 128x128 px
mean: 128x128 px
max: 128x128 px
0: ~79.81%
1: ~20.19%

Samples:

text	image	label
`a cobain glasses character with a gradient 2 head and purple puffballs hair wearing a white sweater, gradient 4 background`		`1`
`a cobain glasses character with a gradient 2 head and purple puffballs hair wearing a white sweater, gradient 4 background`		`0`
`a cobain glasses character with a gradient 2 head and purple puffballs hair wearing a white sweater, gradient 4 background`		`0`

Loss: BinaryCrossEntropyLoss with these parameters:
json
1{ 2 "activation_fn": "torch.nn.modules.linear.Identity", 3 "pos_weight": null 4}

Evaluation Datasets

image_to_text

Dataset: image_to_text at a575ac6
Size: 500 evaluation samples
Columns: image, text, and label
Approximate statistics based on the first 100 samples:
image text label
type image string int
modality image text
details
min: 128x128 px
mean: 128x128 px
max: 128x128 px
min: 27 tokens
mean: 32.94 tokens
max: 37 tokens
0: ~79.81%
1: ~20.19%

Samples:

image	text	label
	`a content character with a tan head and purple puffballs hair wearing a blue fleece, green background`	`1`
	`a mustache character with a orange head and yellow headband hair wearing a white turtleneck, gradient 3 background`	`0`
	`a sunglasses character with a purple head and yellow bowlcut hair wearing a orange collar, gradient 4 background`	`0`

Loss: BinaryCrossEntropyLoss with these parameters:
json
1{ 2 "activation_fn": "torch.nn.modules.linear.Identity", 3 "pos_weight": null 4}

text_to_image

Dataset: text_to_image at a575ac6
Size: 500 evaluation samples
Columns: text, image, and label
Approximate statistics based on the first 100 samples:
text image label
type string image int
modality text image
details
min: 29 tokens
mean: 32.73 tokens
max: 36 tokens
min: 128x128 px
mean: 128x128 px
max: 128x128 px
0: ~79.81%
1: ~20.19%

Samples:

text	image	label
`a content character with a tan head and purple puffballs hair wearing a blue fleece, green background`		`1`
`a content character with a tan head and purple puffballs hair wearing a blue fleece, green background`		`0`
`a content character with a tan head and purple puffballs hair wearing a blue fleece, green background`		`0`

Loss: BinaryCrossEntropyLoss with these parameters:
json
1{ 2 "activation_fn": "torch.nn.modules.linear.Identity", 3 "pos_weight": null 4}

Training Hyperparameters

Non-Default Hyperparameters

num_train_epochs: 1
learning_rate: 5e-06
warmup_steps: 0.1
fp16: True
prompts: {'image_to_text': "Given the image, judge whether the text matches it. Respond with 1 if they match, 0 if they don't.", 'text_to_image': "Given the text, judge whether the image matches it. Respond with 1 if they match, 0 if they don't."}

All Hyperparameters

Click to expand

per_device_train_batch_size: 8
num_train_epochs: 1
max_steps: -1
learning_rate: 5e-06
lr_scheduler_type: linear
lr_scheduler_kwargs: None
warmup_steps: 0.1
optim: adamw_torch_fused
optim_args: None
weight_decay: 0.0
adam_beta1: 0.9
adam_beta2: 0.999
adam_epsilon: 1e-08
optim_target_modules: None
gradient_accumulation_steps: 1
average_tokens_across_devices: True
max_grad_norm: 1.0
label_smoothing_factor: 0.0
bf16: False
fp16: True
bf16_full_eval: False
fp16_full_eval: False
tf32: None
gradient_checkpointing: False
gradient_checkpointing_kwargs: None
torch_compile: False
torch_compile_backend: None
torch_compile_mode: None
use_liger_kernel: False
liger_kernel_config: None
use_cache: False
neftune_noise_alpha: None
torch_empty_cache_steps: None
auto_find_batch_size: False
log_on_each_node: True
logging_nan_inf_filter: True
include_num_input_tokens_seen: no
log_level: passive
log_level_replica: warning
disable_tqdm: False
project: huggingface
trackio_space_id: None
trackio_bucket_id: None
trackio_static_space_id: None
per_device_eval_batch_size: 8
prediction_loss_only: True
eval_on_start: False
eval_do_concat_batches: True
eval_use_gather_object: False
eval_accumulation_steps: None
include_for_metrics: []
batch_eval_metrics: False
save_only_model: False
save_on_each_node: False
enable_jit_checkpoint: False
push_to_hub: False
hub_private_repo: None
hub_model_id: None
hub_strategy: every_save
hub_always_push: False
hub_revision: None
load_best_model_at_end: False
ignore_data_skip: False
restore_callback_states_from_checkpoint: False
full_determinism: False
seed: 42
data_seed: None
use_cpu: False
accelerator_config: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}
parallelism_config: None
dataloader_drop_last: False
dataloader_num_workers: 0
dataloader_pin_memory: True
dataloader_persistent_workers: False
dataloader_prefetch_factor: None
remove_unused_columns: True
label_names: None
train_sampling_strategy: random
length_column_name: length
ddp_find_unused_parameters: None
ddp_bucket_cap_mb: None
ddp_broadcast_buffers: False
ddp_static_graph: None
ddp_backend: None
ddp_timeout: 1800
fsdp: None
fsdp_config: None
deepspeed: None
debug: []
skip_memory_metrics: True
do_predict: False
resume_from_checkpoint: None
warmup_ratio: None
local_rank: -1
prompts: {'image_to_text': "Given the image, judge whether the text matches it. Respond with 1 if they match, 0 if they don't.", 'text_to_image': "Given the text, judge whether the image matches it. Respond with 1 if they match, 0 if they don't."}
batch_sampler: batch_sampler
multi_dataset_batch_sampler: proportional
router_mapping: {}
learning_rate_mapping: {}

Training Logs

Epoch	Step	Training Loss	image to text loss	text to image loss	doodles-image-to-text-eval_ndcg@10	doodles-text-to-image-eval_ndcg@10
0.1004	113	0.3237	-	-	-	-
0.2007	226	0.3331	-	-	-	-
0.2504	282	-	0.2869	0.2543	0.9037	0.6832
0.3011	339	0.2455	-	-	-	-
0.4014	452	0.2192	-	-	-	-
0.5009	564	-	0.1758	0.2778	0.9704	0.7909
0.5018	565	0.1695	-	-	-	-
0.6021	678	0.1765	-	-	-	-
0.7025	791	0.2051	-	-	-	-
0.7513	846	-	0.1583	0.2085	0.9869	0.8108
0.8028	904	0.1758	-	-	-	-
0.9032	1017	0.2113	-	-	-	-
1.0	1126	-	0.1462	0.1767	0.9869	0.8167
-1	-1	-	-	-	0.9869	0.8167

Training Time

Training: 2.0 hours

Framework Versions

Python: 3.12.13
Sentence Transformers: 5.6.0
Transformers: 5.13.1
PyTorch: 2.11.0+cu128
Accelerate: 1.14.0
Datasets: 4.0.0
Tokenizers: 0.22.2

Additional Resources

Training and Finetuning Reranker Models with Sentence Transformers: the end-to-end guide for training or finetuning Cross Encoder (reranker) models.
Multimodal Embedding & Reranker Models with Sentence Transformers: use text, image, audio, and video reranker models through the same API.
Training and Finetuning Multimodal Embedding & Reranker Models with Sentence Transformers: training multimodal Cross Encoders.

Citation

BibTeX

Sentence Transformers

bibtex

@inproceedings{reimers-2019-sentence-bert,
    title = "Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks",
    author = "Reimers, Nils and Gurevych, Iryna",
    booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing",
    month = "11",
    year = "2019",
    publisher = "Association for Computational Linguistics",
    url = "https://arxiv.org/abs/1908.10084",
}

	image	text	label
type	image	string	int
modality	image	text
details	min: 128x128 px mean: 128x128 px max: 128x128 px	min: 30 tokens mean: 33.48 tokens max: 37 tokens	0: ~79.81% 1: ~20.19%