Amrlib

model_parse_xfm_bart_large-v0_1_0.tar.gz md5sum: 5633e5e6f8a4f4398ce64abab71c473a
MetaData (amrlib_meta.json)
{
"model_type":"stog",
"version":"0.1.0",
"date":"2022-02-16",
"inference_module":".parse_xfm.inference",
"inference_class":"Inference",
"model_fn":"pytorch_model.bin",
"base_model":"facebook/bart-large",
"kwargs":{}
}

parse_xfm_bart_base-v0_1_0
Sentence to Graph parse model based on the pre-trained HuggingFace transformer model: 'facebook/bart-base'

model_parse_xfm_bart_base-v0_1_0.tar.gz md5sum: f8eb889468f4d0c3c0677fc910fb2240
MetaData (amrlib_meta.json)
{
"model_type":"stog",
"version":"0.1.0",
"date":"2022-02-16",
"inference_module":".parse_xfm.inference",
"inference_class":"Inference",
"model_fn":"pytorch_model.bin",
"base_model":"facebook/bart-base",
"kwargs":{}
}


model_parse_t5-v0_2_0
Sentence to Graph parse model based on the pre-trained HuggingFace T5 transformer.
SMATCH score = 82.

model_parse_t5-v0_2_0.tar.gz md5sum: b5bbd010c79b87072d1ad4923910c0d7

MetaData (amrlib_meta.json)
{
"model_type":"stog",
"version":"0.2.0",
"date":"2021-11-27",
"inference_module":".parse_t5.inference",
"inference_class":"Inference",
"model_fn":"pytorch_model.bin",
"kwargs":{}
}


model_parse_spring-v0_1_0
model_parse_spring-v0_1_0

**Sentence to Graph parse model derived from the [SPRING model's code](https://github.com/SapienzaNLP/spring).
SMATCH score = 83.

model_parse_spring-v0_1_0.tar.gz md5sum: 73df04968c5fab39d248fb2406648a0c

MetaData (amrlib_meta.json)
{
"model_type":"stog",
"version":"0.1.0",
"date":"2021-11-25",
"inference_module":".parse_spring.inference",
"inference_class":"Inference",
"model_fn":"model.pt",
"kwargs":{}
}


model_generate_t5wtense-v0_1_0
model_generate_t5wtense-v0_1_0

**Graph to Sentence model based on the pre-trained HuggingFace T5 transformer with tense information**
BLEU score = 44 for basic AMR graphs
BLEU score = 54 with tense information (part of speech tags) added

model_generate_t5wtense-v0_1_0.tar.gz checksum: ef1074064e64062af6b968d046d1dacf

MetaData (amrlib_meta.json)
{
"model_type":"gtos",
"version":"0.1.0",
"date":"2020-12-30",
"inference_module":".generate_t5wtense.inference",
"inference_class":"Inference",
"model_fn":"pytorch_model.bin",
"kwargs":{}
}


model_parse_t5-v0_1_0
model_parse_t5-v0_1_0

**Sentence to Graph parse model based on the pre-trained HuggingFace T5 transformer.**
SMATCH score = 81.

model_parse_t5-v0_1_0.tar.gz checksum: f2da07f2b08a5b4780442bbd3d165f02

MetaData (amrlib_meta.json)
{
"model_type":"stog",
"version":"0.1.0",
"date":"2020-12-12",
"inference_module":".parse_t5.inference",
"inference_class":"Inference",
"model_fn":"pytorch_model.bin",
"kwargs":{}
}

* Add generate_xfm and removed generate_t5/generate_t5wtense
* Fix parse_gsii annotator multiprocessing error under Windows and Mac
* Add "quiet" option to the parse_gsii model

* Fix missing resources directory for the parse_spring model in the pypi package
- No code changes except setup scripts.

* Add parse_xfm model, configs and training scripts (this replaces parse_t5)
* Add WikiAdderBlink and update parse_xmf / parse_spring train code to use it.
* Added tensorboard smatch logging for parse_xfm training
* Added smatch log redirection to utils/logging.py
* Fix interrogative', 'imperative', 'expressive' as possible nodes

* Added model_parse_spring
* Updated code for model_parse_t5 (faster training and inference)
* Fix dynamic load of model_stog when running amr_view (for transformers 4.4.2)
* Retest with torch 1.10.0 and transformers 4.12.3
* Add additional tests for parse models