This is a Cross Encoder model finetuned from answerdotai/answerai-colbert-small-v1 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: answerdotai/answerai-colbert-small-v1
Maximum Sequence Length: 512 tokens
Number of Output Labels: 1 label

Language: en
License: apache-2.0

Model Sources

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

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("gmguarino/answerai-colbert-small-v1-climatecheck-chunks")
# Get scores for pairs of texts
pairs = [
    ['Turns out, species that can adapt easily to different environments are often the ones that can survive in a wide variety of places. Interesting, right?', 'Welcome to the city Human populations are shifting en masse to cities, which is leading to rapid increases in the number and extent of urban areas. Such changes are well known to cause declines in many species, but they can also act as alternative selection pressures to which some species are able to adapt. Johnson and Munshi-South review the suite of pressures that urban environments exert, the ways in which species may (or may not) adapt, and the larger impact of these evolutionary events on natural processes and human populations. Understanding such urban evolution patterns will improve our ability to foster species persistence in the face of urbanization and to mitigate some of the challenges, such as disease, that adaptation can bring.'],
    ['Turns out, species that can adapt easily to different environments are often the ones that can survive in a wide variety of places. Interesting, right?', 'Science, this issue p. eaam8327 BACKGROUND The extent of urban areas is increasing around the world, and most humans now live in cities. Urbanization results in dramatic environmental change, including increased temperatures, more impervious surface cover, altered hydrology, and elevated pollution. Urban areas also host more non-native species and reduced abundance and diversity of many native species. These environmental changes brought by global urbanization are creating novel ecosystems with unknown consequences for the evolution of life. Here, we consider how early human settlements led to the evolution of human commensals, including some of the most notorious pests and disease vectors. We also comprehensively review how contemporary urbanization affects the evolution of species that coinhabit cities.'],
    ['Turns out, species that can adapt easily to different environments are often the ones that can survive in a wide variety of places. Interesting, right?', 'We also comprehensively review how contemporary urbanization affects the evolution of species that coinhabit cities. ADVANCES A recent surge of research shows that urbanization affects both nonadaptive and adaptive evolution. Some of the clearest results of urban evolution show that cities elevate the strength of random genetic drift (stochastic changes in allele frequencies) and restrict gene flow (the movement of alleles between populations due to dispersal and mating). Populations of native species in cities often represent either relicts that predate urbanization or populations that established after a city formed. Both scenarios frequently result in a loss of genetic diversity within populations and increased differentiation between populations.'],
    ['Turns out, species that can adapt easily to different environments are often the ones that can survive in a wide variety of places. Interesting, right?', 'Both scenarios frequently result in a loss of genetic diversity within populations and increased differentiation between populations. Fragmentation and urban infrastructure also create barriers to dispersal, and consequently, gene flow is often reduced among city populations, which further contributes to genetic differentiation between populations. The influence of urbanization on mutation and adaptive evolution are less clear. A small number of studies suggest that industrial pollution can elevate mutation rates, but the pervasiveness of this effect is unknown. A better studied phenomenon are the effects of urbanization on evolution by natural selection. A growing number of studies show that plant and animal populations experience divergent selection between urban and nonurban environments.'],
    ['Turns out, species that can adapt easily to different environments are often the ones that can survive in a wide variety of places. Interesting, right?', 'A growing number of studies show that plant and animal populations experience divergent selection between urban and nonurban environments. This divergent selection has led to adaptive evolution in life history, morphology, physiology, behavior, and reproductive traits. These adaptations typically evolve in response to pesticide use, pollution, local climate, or the physical structure of cities. Despite these important results, the genetic basis of adaptive evolution is known from only a few cases. Most studies also examine only a few populations in one city, and experimental validation is rare. OUTLOOK The study of evolution in urban areas provides insights into both fundamental and applied problems in biology.'],
]
scores = model.predict(pairs)
print(scores.shape)
# (5,)

# Or rank different texts based on similarity to a single text
ranks = model.rank(
    'Turns out, species that can adapt easily to different environments are often the ones that can survive in a wide variety of places. Interesting, right?',
    [
        'Welcome to the city Human populations are shifting en masse to cities, which is leading to rapid increases in the number and extent of urban areas. Such changes are well known to cause declines in many species, but they can also act as alternative selection pressures to which some species are able to adapt. Johnson and Munshi-South review the suite of pressures that urban environments exert, the ways in which species may (or may not) adapt, and the larger impact of these evolutionary events on natural processes and human populations. Understanding such urban evolution patterns will improve our ability to foster species persistence in the face of urbanization and to mitigate some of the challenges, such as disease, that adaptation can bring.',
        'Science, this issue p. eaam8327 BACKGROUND The extent of urban areas is increasing around the world, and most humans now live in cities. Urbanization results in dramatic environmental change, including increased temperatures, more impervious surface cover, altered hydrology, and elevated pollution. Urban areas also host more non-native species and reduced abundance and diversity of many native species. These environmental changes brought by global urbanization are creating novel ecosystems with unknown consequences for the evolution of life. Here, we consider how early human settlements led to the evolution of human commensals, including some of the most notorious pests and disease vectors. We also comprehensively review how contemporary urbanization affects the evolution of species that coinhabit cities.',
        'We also comprehensively review how contemporary urbanization affects the evolution of species that coinhabit cities. ADVANCES A recent surge of research shows that urbanization affects both nonadaptive and adaptive evolution. Some of the clearest results of urban evolution show that cities elevate the strength of random genetic drift (stochastic changes in allele frequencies) and restrict gene flow (the movement of alleles between populations due to dispersal and mating). Populations of native species in cities often represent either relicts that predate urbanization or populations that established after a city formed. Both scenarios frequently result in a loss of genetic diversity within populations and increased differentiation between populations.',
        'Both scenarios frequently result in a loss of genetic diversity within populations and increased differentiation between populations. Fragmentation and urban infrastructure also create barriers to dispersal, and consequently, gene flow is often reduced among city populations, which further contributes to genetic differentiation between populations. The influence of urbanization on mutation and adaptive evolution are less clear. A small number of studies suggest that industrial pollution can elevate mutation rates, but the pervasiveness of this effect is unknown. A better studied phenomenon are the effects of urbanization on evolution by natural selection. A growing number of studies show that plant and animal populations experience divergent selection between urban and nonurban environments.',
        'A growing number of studies show that plant and animal populations experience divergent selection between urban and nonurban environments. This divergent selection has led to adaptive evolution in life history, morphology, physiology, behavior, and reproductive traits. These adaptations typically evolve in response to pesticide use, pollution, local climate, or the physical structure of cities. Despite these important results, the genetic basis of adaptive evolution is known from only a few cases. Most studies also examine only a few populations in one city, and experimental validation is rare. OUTLOOK The study of evolution in urban areas provides insights into both fundamental and applied problems in biology.',
    ]
)
# [{'corpus_id': ..., 'score': ...}, {'corpus_id': ..., 'score': ...}, ...]

Training Details

Training Dataset

Unnamed Dataset

Size: 3,952 training samples
Columns: anchor, passage, and label
Approximate statistics based on the first 1000 samples:
anchor passage label
type string string float
details
min: 30 characters
mean: 107.43 characters
max: 209 characters
min: 83 characters
mean: 598.79 characters
max: 954 characters
min: 0.0
mean: 0.53
max: 1.0

Samples:

anchor	passage	label
`Turns out, species that can adapt easily to different environments are often the ones that can survive in a wide variety of places. Interesting, right?`	Welcome to the city Human populations are shifting en masse to cities, which is leading to rapid increases in the number and extent of urban areas. Such changes are well known to cause declines in many species, but they can also act as alternative selection pressures to which some species are able to adapt. Johnson and Munshi-South review the suite of pressures that urban environments exert, the ways in which species may (or may not) adapt, and the larger impact of these evolutionary events on natural processes and human populations. Understanding such urban evolution patterns will improve our ability to foster species persistence in the face of urbanization and to mitigate some of the challenges, such as disease, that adaptation can bring.	`0.0`
`Turns out, species that can adapt easily to different environments are often the ones that can survive in a wide variety of places. Interesting, right?`	Science, this issue p. eaam8327 BACKGROUND The extent of urban areas is increasing around the world, and most humans now live in cities. Urbanization results in dramatic environmental change, including increased temperatures, more impervious surface cover, altered hydrology, and elevated pollution. Urban areas also host more non-native species and reduced abundance and diversity of many native species. These environmental changes brought by global urbanization are creating novel ecosystems with unknown consequences for the evolution of life. Here, we consider how early human settlements led to the evolution of human commensals, including some of the most notorious pests and disease vectors. We also comprehensively review how contemporary urbanization affects the evolution of species that coinhabit cities.	`0.0`
`Turns out, species that can adapt easily to different environments are often the ones that can survive in a wide variety of places. Interesting, right?`	We also comprehensively review how contemporary urbanization affects the evolution of species that coinhabit cities. ADVANCES A recent surge of research shows that urbanization affects both nonadaptive and adaptive evolution. Some of the clearest results of urban evolution show that cities elevate the strength of random genetic drift (stochastic changes in allele frequencies) and restrict gene flow (the movement of alleles between populations due to dispersal and mating). Populations of native species in cities often represent either relicts that predate urbanization or populations that established after a city formed. Both scenarios frequently result in a loss of genetic diversity within populations and increased differentiation between populations.	`0.0`

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

per_device_train_batch_size: 16
learning_rate: 2e-05
num_train_epochs: 5
warmup_ratio: 0.1
fp16: True
dataloader_num_workers: 2

All Hyperparameters

Click to expand

overwrite_output_dir: False
do_predict: False
eval_strategy: no
prediction_loss_only: True
per_device_train_batch_size: 16
per_device_eval_batch_size: 8
per_gpu_train_batch_size: None
per_gpu_eval_batch_size: None
gradient_accumulation_steps: 1
eval_accumulation_steps: None
torch_empty_cache_steps: None
learning_rate: 2e-05
weight_decay: 0.0
adam_beta1: 0.9
adam_beta2: 0.999
adam_epsilon: 1e-08
max_grad_norm: 1.0
num_train_epochs: 5
max_steps: -1
lr_scheduler_type: linear
lr_scheduler_kwargs: {}
warmup_ratio: 0.1
warmup_steps: 0
log_level: passive
log_level_replica: warning
log_on_each_node: True
logging_nan_inf_filter: True
save_safetensors: True
save_on_each_node: False
save_only_model: False
restore_callback_states_from_checkpoint: False
no_cuda: False
use_cpu: False
use_mps_device: False
seed: 42
data_seed: None
jit_mode_eval: False
use_ipex: False
bf16: False
fp16: True
fp16_opt_level: O1
half_precision_backend: auto
bf16_full_eval: False
fp16_full_eval: False
tf32: None
local_rank: 0
ddp_backend: None
tpu_num_cores: None
tpu_metrics_debug: False
debug: []
dataloader_drop_last: False
dataloader_num_workers: 2
dataloader_prefetch_factor: None
past_index: -1
disable_tqdm: False
remove_unused_columns: True
label_names: None
load_best_model_at_end: False
ignore_data_skip: False
fsdp: []
fsdp_min_num_params: 0
fsdp_config: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
tp_size: 0
fsdp_transformer_layer_cls_to_wrap: None
accelerator_config: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}
deepspeed: None
label_smoothing_factor: 0.0
optim: adamw_torch
optim_args: None
adafactor: False
group_by_length: False
length_column_name: length
ddp_find_unused_parameters: None
ddp_bucket_cap_mb: None
ddp_broadcast_buffers: False
dataloader_pin_memory: True
dataloader_persistent_workers: False
skip_memory_metrics: True
use_legacy_prediction_loop: False
push_to_hub: False
resume_from_checkpoint: None
hub_model_id: None
hub_strategy: every_save
hub_private_repo: None
hub_always_push: False
gradient_checkpointing: False
gradient_checkpointing_kwargs: None
include_inputs_for_metrics: False
include_for_metrics: []
eval_do_concat_batches: True
fp16_backend: auto
push_to_hub_model_id: None
push_to_hub_organization: None
mp_parameters:
auto_find_batch_size: False
full_determinism: False
torchdynamo: None
ray_scope: last
ddp_timeout: 1800
torch_compile: False
torch_compile_backend: None
torch_compile_mode: None
include_tokens_per_second: False
include_num_input_tokens_seen: False
neftune_noise_alpha: None
optim_target_modules: None
batch_eval_metrics: False
eval_on_start: False
use_liger_kernel: False
eval_use_gather_object: False
average_tokens_across_devices: False
prompts: None
batch_sampler: batch_sampler
multi_dataset_batch_sampler: proportional

Training Logs

Click to expand

Epoch	Step	Training Loss
0.0040	1	0.6585
0.0405	10	0.6842
0.0810	20	0.6886
0.1215	30	0.6834
0.1619	40	0.69
0.2024	50	0.6837
0.2429	60	0.685
0.0040	1	0.6433
0.0202	5	0.6869
0.0405	10	0.6759
0.0607	15	0.6893
0.0810	20	0.6853
0.1012	25	0.6718
0.1215	30	0.6883
0.1417	35	0.6805
0.1619	40	0.698
0.1822	45	0.6801
0.2024	50	0.682
0.2227	55	0.676
0.2429	60	0.6905
0.2632	65	0.6841
0.2834	70	0.6837
0.0040	1	0.6265
0.0202	5	0.6856
0.0405	10	0.6707
0.0607	15	0.6894
0.0810	20	0.6832
0.1012	25	0.665
0.1215	30	0.6877
0.1417	35	0.6769
0.1619	40	0.7012
0.1822	45	0.6763
0.2024	50	0.679
0.2227	55	0.6705
0.2429	60	0.6917
0.2632	65	0.6818
0.2834	70	0.6821
0.3036	75	0.6588
0.3239	80	0.6492
0.3441	85	0.6676
0.3644	90	0.6397
0.3846	95	0.666
0.4049	100	0.696
0.4251	105	0.683
0.4453	110	0.6527
0.4656	115	0.7002
0.4858	120	0.7081
0.5061	125	0.7401
0.5263	130	0.6776
0.5466	135	0.6653
0.5668	140	0.6787
0.5870	145	0.6678
0.6073	150	0.6821
0.6275	155	0.6536
0.6478	160	0.6869
0.6680	165	0.6754
0.6883	170	0.7247
0.7085	175	0.7043
0.7287	180	0.6749
0.7490	185	0.6787
0.7692	190	0.6628
0.7895	195	0.6699
0.8097	200	0.6463
0.8300	205	0.6361
0.8502	210	0.619
0.8704	215	0.6158
0.8907	220	0.6604
0.9109	225	0.5839
0.9312	230	0.5836
0.9514	235	0.5937
0.9717	240	0.6356
0.9919	245	0.5775
1.0121	250	0.5566
1.0324	255	0.5879
1.0526	260	0.5725
1.0729	265	0.5549
1.0931	270	0.5182
1.1134	275	0.5257
1.1336	280	0.5633
1.1538	285	0.5348
1.1741	290	0.5485
1.1943	295	0.574
1.2146	300	0.5967
1.2348	305	0.4148
1.2551	310	0.5355
1.2753	315	0.5388
1.2955	320	0.4969
1.3158	325	0.4887
1.3360	330	0.5494
1.3563	335	0.4695
1.3765	340	0.6148
1.3968	345	0.6179
1.4170	350	0.555
1.4372	355	0.4942
1.4575	360	0.4936
1.4777	365	0.4824
1.4980	370	0.4783
1.5182	375	0.6457
1.5385	380	0.4025
1.5587	385	0.4587
1.5789	390	0.5683
1.5992	395	0.5296
1.6194	400	0.4801
1.6397	405	0.452
1.6599	410	0.3888
1.6802	415	0.4634
1.7004	420	0.5594
1.7206	425	0.4489
1.7409	430	0.5764
1.7611	435	0.4233
1.7814	440	0.4016
1.8016	445	0.4774
1.8219	450	0.5146
1.8421	455	0.601
1.8623	460	0.4857
1.8826	465	0.4385
1.9028	470	0.474
1.9231	475	0.4027
1.9433	480	0.557
1.9636	485	0.5921
1.9838	490	0.4424
2.0040	495	0.5546
2.0243	500	0.4989
2.0445	505	0.5128
2.0648	510	0.4469
2.0850	515	0.4591
2.1053	520	0.4151
2.1255	525	0.5473
2.1457	530	0.4153
2.1660	535	0.3811
2.1862	540	0.3508
2.2065	545	0.4734
2.2267	550	0.3578
2.2470	555	0.3539
2.2672	560	0.3924
2.2874	565	0.3067
2.3077	570	0.3795
2.3279	575	0.37
2.3482	580	0.3612
2.3684	585	0.3223
2.3887	590	0.4666
2.4089	595	0.4536
2.4291	600	0.4246
2.4494	605	0.4609
2.4696	610	0.404
2.4899	615	0.4847
2.5101	620	0.5884
2.5304	625	0.5785
2.5506	630	0.5211
2.5709	635	0.3566
2.5911	640	0.3911
2.6113	645	0.4295
2.6316	650	0.3605
2.6518	655	0.5329
2.6721	660	0.4455
2.6923	665	0.3665
2.7126	670	0.4392
2.7328	675	0.3559
2.7530	680	0.4053
2.7733	685	0.5254
2.7935	690	0.4304
2.8138	695	0.3854
2.8340	700	0.4575
2.8543	705	0.5575
2.8745	710	0.4285
2.8947	715	0.4409
2.9150	720	0.3567
2.9352	725	0.3528
2.9555	730	0.4723
2.9757	735	0.4635
2.9960	740	0.4086
3.0162	745	0.3767
3.0364	750	0.4548
3.0567	755	0.3311
3.0769	760	0.3817
3.0972	765	0.3231
3.1174	770	0.479
3.1377	775	0.4462
3.1579	780	0.383
3.1781	785	0.3326
3.1984	790	0.4318
3.2186	795	0.4167
3.2389	800	0.487
3.2591	805	0.362
3.2794	810	0.3862
3.2996	815	0.4245
3.3198	820	0.261
3.3401	825	0.3718
3.3603	830	0.3077
3.3806	835	0.3098
3.4008	840	0.2903
3.4211	845	0.4764
3.4413	850	0.3676
3.4615	855	0.4486
3.4818	860	0.3227
3.5020	865	0.3489
3.5223	870	0.4432
3.5425	875	0.3406
3.5628	880	0.4052
3.5830	885	0.2647
3.6032	890	0.399
3.6235	895	0.2908
3.6437	900	0.4351
3.6640	905	0.3273
3.6842	910	0.4671
3.7045	915	0.2794
3.7247	920	0.4279
3.7449	925	0.239
3.7652	930	0.3938
3.7854	935	0.4376
3.8057	940	0.4792
3.8259	945	0.3866
3.8462	950	0.2753
3.8664	955	0.2502
3.8866	960	0.3265
3.9069	965	0.4292
3.9271	970	0.2987
3.9474	975	0.3569
3.9676	980	0.3146
3.9879	985	0.3535
4.0081	990	0.4034
4.0283	995	0.4426
4.0486	1000	0.3035
4.0688	1005	0.2956
4.0891	1010	0.3476
4.1093	1015	0.3306
4.1296	1020	0.2093
4.1498	1025	0.3127
4.1700	1030	0.3995
4.1903	1035	0.4798
4.2105	1040	0.2275
4.2308	1045	0.4751
4.2510	1050	0.4534
4.2713	1055	0.3419
4.2915	1060	0.3475
4.3117	1065	0.2916
4.3320	1070	0.3456
4.3522	1075	0.3619
4.3725	1080	0.2504
4.3927	1085	0.2638
4.4130	1090	0.3414
4.4332	1095	0.2609
4.4534	1100	0.2555
4.4737	1105	0.3007
4.4939	1110	0.3586
4.5142	1115	0.4047
4.5344	1120	0.271
4.5547	1125	0.2517
4.5749	1130	0.4167
4.5951	1135	0.341
4.6154	1140	0.3734
4.6356	1145	0.3632
4.6559	1150	0.4568
4.6761	1155	0.3237
4.6964	1160	0.4222
4.7166	1165	0.2528
4.7368	1170	0.2831
4.7571	1175	0.5008
4.7773	1180	0.2495
4.7976	1185	0.3158
4.8178	1190	0.5574
4.8381	1195	0.3171
4.8583	1200	0.193
4.8785	1205	0.408
4.8988	1210	0.2704
4.9190	1215	0.2975
4.9393	1220	0.1966
4.9595	1225	0.38
4.9798	1230	0.4851
5.0	1235	0.3162

Framework Versions

Python: 3.11.12
Sentence Transformers: 4.1.0
Transformers: 4.51.3
PyTorch: 2.6.0+cu124
Accelerate: 1.6.0
Datasets: 3.6.0
Tokenizers: 0.21.1

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",
}

	anchor	passage	label
type	string	string	float
details	min: 30 characters mean: 107.43 characters max: 209 characters	min: 83 characters mean: 598.79 characters max: 954 characters	min: 0.0 mean: 0.53 max: 1.0