ProtSent-V2.5 150M: V2 + GOR + DMS CoSENT

1_Pooling/config.json

@@ -0,0 +1,5 @@
+{
+    "embedding_dimension": 640,
+    "pooling_mode": "mean",
+    "include_prompt": true
+}

config.json

@@ -0,0 +1,44 @@
+{
+  "add_pooling_layer": false,
+  "architectures": [
+    "EsmModel"
+  ],
+  "attention_probs_dropout_prob": 0.0,
+  "attn_backend": "sdpa",
+  "bos_token_id": 0,
+  "classifier_dropout": null,
+  "dtype": "float32",
+  "emb_layer_norm_before": false,
+  "eos_token_id": 2,
+  "esmfold_config": null,
+  "fastplms_checkpoint_hash": "c1d1ed6ec8cad4600e8b89bc023343fb9a67aa6b574cfbf6f46189c15478c159",
+  "fastplms_checkpoint_repo_id": "Synthyra/ESM2-150M",
+  "fastplms_checkpoint_revision": "979e0880dfc9e0c0080839b83d9d2dc05b92786a",
+  "fastplms_model_id": "esm2_150m",
+  "fastplms_release_tool_revision": "e6dd397a9ad368c998d714f6bd64d40b533d1ed1",
+  "fastplms_release_tool_sha256": "6d335c05aa49a232086a816deb25d248d1490529e5783acc3355b9bc6f03e0c2",
+  "fastplms_runtime_bundle_sha256": "f22353ef386f607784889dc303488a2dc5e99b009a5e423ab3eff1f1eb03a9a4",
+  "fastplms_runtime_revision": "e6dd397a9ad368c998d714f6bd64d40b533d1ed1",
+  "fastplms_source_tree_sha256": "818bf2bf5f573adc842b55f4f8feff926ddc25f904c5769ff5c382077d3a78a2",
+  "fastplms_weights_revision": "979e0880dfc9e0c0080839b83d9d2dc05b92786a",
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.0,
+  "hidden_size": 640,
+  "initializer_range": 0.02,
+  "intermediate_size": 2560,
+  "is_folding_model": false,
+  "layer_norm_eps": 1e-05,
+  "mask_token_id": 32,
+  "max_position_embeddings": 1026,
+  "model_type": "esm",
+  "num_attention_heads": 20,
+  "num_hidden_layers": 30,
+  "pad_token_id": 1,
+  "position_embedding_type": "rotary",
+  "tie_word_embeddings": false,
+  "token_dropout": true,
+  "transformers_version": "5.14.1",
+  "use_cache": false,
+  "vocab_list": null,
+  "vocab_size": 33
+}

config_sentence_transformers.json

@@ -0,0 +1,14 @@
+{
+  "__version__": {
+    "pytorch": "2.13.0+cu130",
+    "sentence_transformers": "5.6.1",
+    "transformers": "5.14.1"
+  },
+  "default_prompt_name": null,
+  "model_type": "SentenceTransformer",
+  "prompts": {
+    "document": "",
+    "query": ""
+  },
+  "similarity_fn_name": "cosine"
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:77b2e675d30e4b39e4b1283804607e68633a4c7cece94d7e951931312f96ff79
+size 592707264

modeling_fastplms.py

@@ -0,0 +1,192 @@
+"""Generated bridge to the embedded FastPLMs runtime sources."""
+
+import base64
+import hashlib
+import importlib
+import importlib.util
+import sys
+import tempfile
+from io import BytesIO
+from pathlib import Path
+from zipfile import ZIP_DEFLATED, ZipFile
+
+from .fastplms_bundle import RUNTIME_DATA, RUNTIME_HASH
+
+if RUNTIME_HASH != "f22353ef386f607784889dc303488a2dc5e99b009a5e423ab3eff1f1eb03a9a4":
+    raise RuntimeError("FastPLMs runtime identity differs from the bridge.")
+
+_RUNTIME_TEMPORARIES = []
+
+def _archive_runtime_hashes(payload):
+    result = {}
+    with ZipFile(BytesIO(payload)) as archive:
+        for member in archive.infolist():
+            name = member.filename
+            parts = Path(name).parts
+            if (
+                member.is_dir()
+                or "\\" in name
+                or not parts
+                or parts[0] != "fastplms"
+                or len(parts) < 2
+                or any(part in {"", ".", ".."} for part in parts)
+                or Path(name).suffix in {".pyc", ".pyo"}
+                or member.flag_bits & 0x1
+                or member.compress_type != ZIP_DEFLATED
+                or member.external_attr >> 16 != 0o100644
+            ):
+                raise RuntimeError("Embedded FastPLMs archive has an unsafe path.")
+            relative = Path(*parts[1:]).as_posix()
+            if relative in result:
+                raise RuntimeError("Embedded FastPLMs archive repeats a path.")
+            result[relative] = hashlib.sha256(archive.read(member)).hexdigest()
+    return result
+
+def _ensure_runtime():
+    payload = base64.b85decode("".join(RUNTIME_DATA))
+    if hashlib.sha256(payload).hexdigest() != RUNTIME_HASH:
+        raise RuntimeError("Embedded FastPLMs runtime hash mismatch.")
+    expected = _archive_runtime_hashes(payload)
+    temporary = tempfile.TemporaryDirectory(prefix="fastplms-artifact-runtime-")
+    try:
+        runtime_root = Path(temporary.name)
+        with ZipFile(BytesIO(payload)) as archive:
+            for member in archive.infolist():
+                target = runtime_root.joinpath(*Path(member.filename).parts)
+                target.parent.mkdir(parents=True, exist_ok=True)
+                with target.open("xb") as handle:
+                    handle.write(archive.read(member))
+        package_root = runtime_root / "fastplms"
+        if _runtime_file_hashes(package_root) != expected:
+            raise RuntimeError(
+                "Private FastPLMs runtime differs from the embedded archive."
+            )
+    except BaseException:
+        temporary.cleanup()
+        raise
+    _RUNTIME_TEMPORARIES.append(temporary)
+    return package_root
+
+def _runtime_file_hashes(package_root):
+    result = {}
+    for path in sorted(package_root.rglob("*")):
+        relative = path.relative_to(package_root)
+        if path.is_symlink():
+            raise RuntimeError("Private FastPLMs runtime contains a symlink.")
+        if path.is_dir():
+            continue
+        if path.suffix in {".pyc", ".pyo"}:
+            raise RuntimeError("Private FastPLMs runtime contains bytecode.")
+        if not path.is_file():
+            raise RuntimeError("Private FastPLMs runtime contains a non-file entry.")
+        result[relative.as_posix()] = hashlib.sha256(path.read_bytes()).hexdigest()
+    return result
+
+def _extend_loaded_package_paths(package_root):
+    for name, module in list(sys.modules.items()):
+        if name != "fastplms" and not name.startswith("fastplms."):
+            continue
+        paths = getattr(module, "__path__", None)
+        if paths is None:
+            continue
+        relative = name.split(".")[1:]
+        candidate = package_root.joinpath(*relative)
+        candidate_text = str(candidate)
+        if candidate.is_dir() and candidate_text not in paths:
+            paths.append(candidate_text)
+
+def _merge_runtime(package, package_root):
+    incoming = _runtime_file_hashes(package_root)
+    known = getattr(package, "__fastplms_artifact_runtime_files__", None)
+    if not isinstance(known, dict):
+        raise RuntimeError(
+            "A non-artifact fastplms module is already loaded. Load the Hub artifact "
+            "in a separate Python process."
+        )
+    conflicts = sorted(
+        relative
+        for relative, digest in incoming.items()
+        if relative in known and known[relative] != digest
+    )
+    if conflicts:
+        raise RuntimeError(
+            "FastPLMs artifacts contain incompatible runtime sources at "
+            + ", ".join(repr(path) for path in conflicts[:5])
+            + ". Load incompatible releases in separate Python processes."
+        )
+    known = dict(known)
+    known.update(incoming)
+    package.__fastplms_artifact_runtime_files__ = known
+    roots = list(getattr(package, "__fastplms_artifact_runtime_roots__", ()))
+    if str(package_root) not in roots:
+        roots.append(str(package_root))
+    package.__fastplms_artifact_runtime_roots__ = tuple(roots)
+    temporaries = list(
+        getattr(package, "__fastplms_artifact_runtime_temporaries__", ())
+    )
+    for temporary in _RUNTIME_TEMPORARIES:
+        if temporary not in temporaries:
+            temporaries.append(temporary)
+    package.__fastplms_artifact_runtime_temporaries__ = tuple(temporaries)
+    hashes = set(getattr(package, "__fastplms_artifact_runtime_hashes__", ()))
+    hashes.add(RUNTIME_HASH)
+    package.__fastplms_artifact_runtime_hashes__ = frozenset(hashes)
+    _extend_loaded_package_paths(package_root)
+    return package
+
+def _import_without_bytecode(module_name):
+    previous = sys.dont_write_bytecode
+    sys.dont_write_bytecode = True
+    try:
+        return importlib.import_module(module_name)
+    finally:
+        sys.dont_write_bytecode = previous
+
+def _install_runtime():
+    package = sys.modules.get("fastplms")
+    hashes = getattr(package, "__fastplms_artifact_runtime_hashes__", ())
+    if RUNTIME_HASH in hashes:
+        return package
+    package_root = _ensure_runtime()
+    if package is not None:
+        return _merge_runtime(package, package_root)
+    spec = importlib.util.spec_from_file_location(
+        "fastplms",
+        package_root / "__init__.py",
+        submodule_search_locations=[str(package_root)],
+    )
+    if spec is None or spec.loader is None:
+        raise ImportError("Unable to load the embedded FastPLMs runtime.")
+    package = importlib.util.module_from_spec(spec)
+    package.__fastplms_artifact_runtime_hash__ = RUNTIME_HASH
+    package.__fastplms_artifact_runtime_hashes__ = frozenset({RUNTIME_HASH})
+    package.__fastplms_artifact_runtime_files__ = _runtime_file_hashes(package_root)
+    package.__fastplms_artifact_runtime_roots__ = (str(package_root),)
+    package.__fastplms_artifact_runtime_temporaries__ = tuple(
+        _RUNTIME_TEMPORARIES
+    )
+    sys.modules["fastplms"] = package
+    previous = sys.dont_write_bytecode
+    sys.dont_write_bytecode = True
+    try:
+        try:
+            spec.loader.exec_module(package)
+        except BaseException:
+            sys.modules.pop("fastplms", None)
+            raise
+    finally:
+        sys.dont_write_bytecode = previous
+    return package
+
+_install_runtime()
+_module_181 = _import_without_bytecode("fastplms.models.esm2.modeling_fastesm")
+FastEsmConfig = _module_181.FastEsmConfig
+FastEsmConfig.__module__ = __name__
+FastEsmForMaskedLM = _module_181.FastEsmForMaskedLM
+FastEsmForMaskedLM.__module__ = __name__
+FastEsmForSequenceClassification = _module_181.FastEsmForSequenceClassification
+FastEsmForSequenceClassification.__module__ = __name__
+FastEsmForTokenClassification = _module_181.FastEsmForTokenClassification
+FastEsmForTokenClassification.__module__ = __name__
+FastEsmModel = _module_181.FastEsmModel
+FastEsmModel.__module__ = __name__

modules.json

@@ -0,0 +1,14 @@
+[
+  {
+    "idx": 0,
+    "name": "0",
+    "path": "",
+    "type": "sentence_transformers.base.modules.transformer.Transformer"
+  },
+  {
+    "idx": 1,
+    "name": "1",
+    "path": "1_Pooling",
+    "type": "sentence_transformers.sentence_transformer.modules.pooling.Pooling"
+  }
+]

sentence_bert_config.json

@@ -0,0 +1,10 @@
+{
+    "transformer_task": "feature-extraction",
+    "modality_config": {
+        "text": {
+            "method": "forward",
+            "method_output_name": "last_hidden_state"
+        }
+    },
+    "module_output_name": "token_embeddings"
+}

tokenizer_config.json

@@ -0,0 +1,55 @@
+{
+  "added_tokens_decoder": {
+    "0": {
+      "content": "<cls>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "<pad>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "<eos>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "3": {
+      "content": "<unk>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "32": {
+      "content": "<mask>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "backend": "custom",
+  "bos_token": "<cls>",
+  "cls_token": "<cls>",
+  "eos_token": "<eos>",
+  "is_local": true,
+  "local_files_only": true,
+  "mask_token": "<mask>",
+  "model_max_length": 512,
+  "pad_token": "<pad>",
+  "tokenizer_class": "EsmTokenizer",
+  "unk_token": "<unk>"
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5c4bffef5750f258ba5a14fbb998f3338069af62e17eb29961449cc2e83b37b5
+size 5585

vocab.txt

@@ -0,0 +1,33 @@
+<cls>
+<pad>
+<eos>
+<unk>
+L
+A
+G
+V
+S
+E
+R
+T
+I
+D
+P
+K
+Q
+N
+F
+Y
+M
+H
+W
+C
+X
+B
+U
+Z
+O
+.
+-
+<null_1>
+<mask>