Add ORB v2 (#18)

* add orbv2 calculator

* add pbe datapoints

* add pbe json file; update pbe error metrics

* bump orb version

mlip_arena/models/externals/orb.py

@@ -41,3 +41,33 @@ class ORB(ORBCalculator):
 
         orbff = pretrained.orb_v1(weights_path=ckpt_path, device=device)
         super().__init__(orbff, device=device, **kwargs)
+
+class ORBv2(ORBCalculator):
+    def __init__(
+        self,
+        checkpoint=REGISTRY["ORBv2"]["checkpoint"],
+        device=None,
+        **kwargs,
+    ):
+        device = device or get_freer_device()
+
+        cache_dir = Path.home() / ".cache" / "orb"
+        cache_dir.mkdir(parents=True, exist_ok=True)
+        ckpt_path = cache_dir / checkpoint
+
+        url = f"https://storage.googleapis.com/orbitalmaterials-public-models/forcefields/{checkpoint}"
+
+        if not ckpt_path.exists():
+            print(f"Downloading ORB model from {url} to {ckpt_path}...")
+            try:
+                response = requests.get(url, stream=True, timeout=120)
+                response.raise_for_status()
+                with open(ckpt_path, "wb") as f:
+                    for chunk in response.iter_content(chunk_size=8192):
+                        f.write(chunk)
+                print("Download completed.")
+            except requests.exceptions.RequestException as e:
+                raise RuntimeError("Failed to download ORB model.") from e
+
+        orbff = pretrained.orb_v2(weights_path=ckpt_path, device=device)
+        super().__init__(orbff, device=device, **kwargs)

mlip_arena/models/registry.yaml

@@ -20,6 +20,7 @@ MACE-MP(M):
   prediction: EFS
   nvt: true
   npt: true
+  license: MIT
 
 CHGNet:
   module: externals
@@ -42,6 +43,7 @@ CHGNet:
   prediction: EFSM
   nvt: true
   npt: true
+  license: BSD-3-Clause
 
 M3GNet:
   module: externals
@@ -63,6 +65,7 @@ M3GNet:
   prediction: EFS
   nvt: true
   npt: true
+  license: BSD-3-Clause
 
 ORB:
   module: externals
@@ -87,6 +90,7 @@ ORB:
   prediction: EFS
   nvt: true
   npt: true
+  license: Apache-2.0
 
 SevenNet:
   module: externals
@@ -110,6 +114,7 @@ SevenNet:
   prediction: EFS
   nvt: true
   npt: true
+  license: GPL-3.0-only
 
 eqV2(OMat):
   module: externals
@@ -132,6 +137,7 @@ eqV2(OMat):
   date: 2024-10-18
   github: https://github.com/FAIR-Chem/fairchem
   doi: https://arxiv.org/abs/2410.12771
+  license: Modified Apache-2.0 (Meta)
 
 
 EquiformerV2(OC22):
@@ -154,6 +160,7 @@ EquiformerV2(OC22):
   prediction: EF
   nvt: true
   npt: false
+  license:
 
 EquiformerV2(OC20):
   module: externals
@@ -196,6 +203,7 @@ eSCN(OC20):
   prediction: EF
   nvt: true
   npt: false
+  license:
 
 MACE-OFF(M):
   module: externals
@@ -217,6 +225,7 @@ MACE-OFF(M):
   prediction: EFS
   nvt: true
   npt: true
+  license: ASL
 
 ALIGNN:
   module: externals
@@ -238,6 +247,7 @@ ALIGNN:
   github: https://github.com/usnistgov/alignn
   doi: https://doi.org/10.1038/s41524-021-00650-1
   date: 2021-11-15
+  license:
 
 DeepMD:
   module: externals
@@ -256,6 +266,29 @@ DeepMD:
   prediction: EFS
   nvt: true
   npt: true
+  license:
+
+ORBv2:
+  module: externals
+  class: ORBv2
+  family: orb
+  package: orb-models==0.4.0
+  checkpoint: orb-v2-20241011.ckpt
+  username: 
+  last-update: 2024-10-29T00:00:00
+  datetime: 2024-10-29T00:00:00 # TODO: Fake datetime
+  datasets:
+    - MPTrj
+    - Alexandria
+  gpu-tasks:
+    - homonuclear-diatomics
+  github: https://github.com/orbital-materials/orb-models
+  doi: 
+  date: 2024-10-15
+  prediction: EFS
+  nvt: true
+  npt: true
+  license: Apache-2.0
 
 ORBv2(MPTrj):
   module: externals
@@ -274,6 +307,7 @@ ORBv2(MPTrj):
   prediction: EFS
   nvt: true
   npt: true
+  license: Apache-2.0
 
 eqV2(MPTrj-S):
   module: externals
@@ -291,4 +325,5 @@ eqV2(MPTrj-S):
   npt: true
   date: 2024-10-18
   github: https://github.com/FAIR-Chem/fairchem
-  doi: https://arxiv.org/abs/2410.12771
+  doi: https://arxiv.org/abs/2410.12771
+  license: Modified Apache-2.0 (Meta)

mlip_arena/tasks/diatomics/alignn/homonuclear-diatomics.json

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e413393463189a25275bd88d14bcc199ad2eb611ae7f334990450e32bfc258d0
-size 2378447
+oid sha256:3f880af442288d79b925e96723fa07c5e31c1d46e2888ed64382c3828c8b6893
+size 2383737

mlip_arena/tasks/diatomics/chgnet/homonuclear-diatomics.json

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:dbe3b30bfe2781a54344e9e4fa3a2f918205ddcfb362a9da15d669ab79ce1df4
-size 1999101
+oid sha256:d3bec21876e97d0841afbebe3b9adc87e61c4b090917f46842ddfcc507b4870e
+size 2003358

mlip_arena/tasks/diatomics/equiformer/homonuclear-diatomics.json

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c3d71886ce5deb7899571dce40dad0952ffdd831c5f80755f9f5985ecccca376
-size 3756332
+oid sha256:fa2354642db212a30b8779e1a9069c58834e690b5d3f8757177676a7236d9cff
+size 3764416

mlip_arena/tasks/diatomics/escn/homonuclear-diatomics.json

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:343a01b13806fdff92de3d21637d44a9ca16f04956a1b4383a9d32e4d618f67a
-size 1955976
+oid sha256:f24fc8f74dc3e9cc77c2099aefe81226764056db3bcb4a9077d42db68ca821d8
+size 1960014

mlip_arena/tasks/diatomics/fairchem/homonuclear-diatomics.json

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:16144d6e7ff5b05d805f37cef43c03ccbc1f27787242da1a12455f62708069a8
-size 2132550
+oid sha256:65bb77b1dca8bddb5ae2ad13aa1e336a30dbefc95953074ccbe077009ab73e08
+size 2136843

mlip_arena/tasks/diatomics/mace-mp/homonuclear-diatomics.json

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6632312f3081afe0d92db9a93304ee73e53eb2847dc6d163a7a00d9716f6f4da
-size 1911488
+oid sha256:f36170ebca7c2c2c876cad7ff9fc4ed5d5043df3da4000089465b31ca2b5544a
+size 1915526

mlip_arena/tasks/diatomics/mace-off/homonuclear-diatomics.json

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8a1e4f34fa03f3a1d3c976ce697cb56964e4fc7ec3985f49666807836bc6f06d
-size 173470
+oid sha256:1202ad5dcf6434cefd25246deeb5785d58cb70ad83bc54577c9f2d711a6b3974
+size 174038

mlip_arena/tasks/diatomics/matgl/homonuclear-diatomics.json

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c257248d2f62bfd09eb334c49405287bf5bc3bf14cfc6ad0a5890425f559f91c
-size 1854330
+oid sha256:217346241aaa0509e95e5f0ec6eb6dafbcc9ef954fd1f3077a778f0ac420f3f2
+size 1858366

mlip_arena/tasks/diatomics/orb/homonuclear-diatomics.json

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:59ba45a8b9d4f6f2d97339077fec2d4cfdccc897a7c93433aa7f5fd2ffebd7b9
-size 2656227
+oid sha256:b6e846339ca128f4e2a0463f057917b51764e36278f43bf218f1717a01f79692
+size 5197369

mlip_arena/tasks/diatomics/sevennet/homonuclear-diatomics.json

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0c9b80395e2cd99fc533a522dbf20eba7576c80e96393fdf8a92fcbe31823c66
-size 1855775
+oid sha256:29542777bafc7425e0de89e9d6fe45341c414da269420658033d6072864ed6e0
+size 1859815

mlip_arena/tasks/diatomics/vasp/homonuclear-diatomics.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b8588dc4d4e203bf8f77f7f1b9736ac3446185084c67a1ad1cfbf0ec4c08e4ed
+size 11785

pyproject.toml

@@ -46,7 +46,7 @@ run = [
     "chgnet==0.3.8",
     "fairchem-core==1.2.0",
     "sevenn==0.9.3.post1",
-    "orb-models==0.3.1",
+    "orb-models==0.4.0",
     "alignn==2024.5.27",
     "prefect>=3.0.4",
     "prefect-dask"
@@ -65,7 +65,7 @@ test = [
     "chgnet==0.3.8",
     "fairchem-core==1.2.0",
     "sevenn==0.9.3.post1",
-    "orb-models==0.3.1",
+    "orb-models==0.4.0",
     "pynanoflann@git+https://github.com/dwastberg/pynanoflann#egg=af434039ae14bedcbb838a7808924d6689274168",
     "alignn==2024.5.27",
     "pytest",

serve/leaderboard.py

@@ -31,6 +31,7 @@ table = pd.DataFrame(
         "Paper",
         "Checkpoint",
         "First Release",
+        "License",
     ]
 )
 
@@ -48,6 +49,7 @@ for model in MODELS:
         "Paper": metadata.get("doi", None) if metadata else None,
         "Checkpoint": metadata.get("checkpoint", None),
         "First Release": metadata.get("date", None),
+        "License": metadata.get("license", None),
     }
     table = pd.concat([table, pd.DataFrame([new_row])], ignore_index=True)
 

serve/ranks/homonuclear-diatomics.py

@@ -1,4 +1,5 @@
 from pathlib import Path
+from ase.data import chemical_symbols
 
 import numpy as np
 import pandas as pd
@@ -29,21 +30,23 @@ for model in valid_models:
     new_row = {
         "Model": model,
         "Conservation deviation [eV/Å]": rows["conservation-deviation"].mean(),
-        "Spearman's coeff. (Energy - repulsion)": rows[
+        "Spearman's coeff. (E: repulsion)": rows[
             "spearman-repulsion-energy"
         ].mean(),
-        "Spearman's coeff. (Force - descending)": rows[
+        "Spearman's coeff. (F: descending)": rows[
             "spearman-descending-force"
         ].mean(),
         "Tortuosity": rows["tortuosity"].mean(),
         "Energy jump [eV]": rows["energy-jump"].mean(),
         "Force flips": rows["force-flip-times"].mean(),
-        "Spearman's coeff. (Energy - attraction)": rows[
+        "Spearman's coeff. (E: attraction)": rows[
             "spearman-attraction-energy"
         ].mean(),
-        "Spearman's coeff. (Force - ascending)": rows[
+        "Spearman's coeff. (F: ascending)": rows[
             "spearman-ascending-force"
         ].mean(),
+        "PBE energy MAE [eV]": rows["pbe-energy-mae"].mean(),
+        "PBE force MAE [eV/Å]": rows["pbe-force-mae"].mean(),
     }
 
     table = pd.concat([table, pd.DataFrame([new_row])], ignore_index=True)
@@ -54,14 +57,20 @@ table.sort_values("Conservation deviation [eV/Å]", ascending=True, inplace=Tru
 table["Rank"] = np.argsort(table["Conservation deviation [eV/Å]"].to_numpy())
 
 table.sort_values(
-    "Spearman's coeff. (Energy - repulsion)", ascending=True, inplace=True
+    "Spearman's coeff. (E: repulsion)", ascending=True, inplace=True
 )
-table["Rank"] += np.argsort(table["Spearman's coeff. (Energy - repulsion)"].to_numpy())
+table["Rank"] += np.argsort(table["Spearman's coeff. (E: repulsion)"].to_numpy())
 
 table.sort_values(
-    "Spearman's coeff. (Force - descending)", ascending=True, inplace=True
+    "Spearman's coeff. (F: descending)", ascending=True, inplace=True
 )
-table["Rank"] += np.argsort(table["Spearman's coeff. (Force - descending)"].to_numpy())
+table["Rank"] += np.argsort(table["Spearman's coeff. (F: descending)"].to_numpy())
+
+table.sort_values("PBE energy MAE [eV]", ascending=True, inplace=True)
+table["Rank"] += np.argsort(table["PBE energy MAE [eV]"].to_numpy())
+
+table.sort_values("PBE force MAE [eV/Å]", ascending=True, inplace=True)
+table["Rank"] += np.argsort(table["PBE force MAE [eV/Å]"].to_numpy())
 
 table.sort_values("Tortuosity", ascending=True, inplace=True)
 table["Rank"] += np.argsort(table["Tortuosity"].to_numpy())
@@ -70,7 +79,7 @@ table.sort_values("Energy jump [eV]", ascending=True, inplace=True)
 table["Rank"] += np.argsort(table["Energy jump [eV]"].to_numpy())
 
 table.sort_values("Force flips", ascending=True, inplace=True)
-table["Rank"] += np.argsort(table["Force flips"].to_numpy())
+table["Rank"] += np.argsort(np.abs(table["Force flips"].to_numpy() - 1))
 
 table["Rank"] += 1
 
@@ -84,13 +93,15 @@ table = table.reindex(
         "Rank",
         "Rank aggr.",
         "Conservation deviation [eV/Å]",
-        "Spearman's coeff. (Energy - repulsion)",
-        "Spearman's coeff. (Force - descending)",
-        "Tortuosity",
+        "PBE energy MAE [eV]",
+        "PBE force MAE [eV/Å]",
+        "Spearman's coeff. (E: repulsion)",
+        "Spearman's coeff. (F: descending)",
         "Energy jump [eV]",
         "Force flips",
-        "Spearman's coeff. (Energy - attraction)",
-        "Spearman's coeff. (Force - ascending)",
+        "Tortuosity",
+        "Spearman's coeff. (E: attraction)",
+        "Spearman's coeff. (F: ascending)",
     ]
 )
 
@@ -103,11 +114,18 @@ s = (
     .background_gradient(
         cmap="Reds",
         subset=[
-            "Spearman's coeff. (Energy - repulsion)",
-            "Spearman's coeff. (Force - descending)",
+            "Spearman's coeff. (E: repulsion)",
+            "Spearman's coeff. (F: descending)",
         ],
         # vmin=-1, vmax=-0.5
     )
+    .background_gradient(
+        cmap="BuPu",
+        subset=[
+            "PBE energy MAE [eV]",
+            "PBE force MAE [eV/Å]",
+        ],
+    )
     .background_gradient(
         cmap="RdPu",
         subset=["Tortuosity", "Energy jump [eV]", "Force flips"],
@@ -117,16 +135,18 @@ s = (
         subset=["Rank", "Rank aggr."],
     )
     .format(
-        "{:.4f}", 
+        "{:.3f}", 
         subset=[
             "Conservation deviation [eV/Å]",
-            "Spearman's coeff. (Energy - repulsion)",
-            "Spearman's coeff. (Force - descending)",
+            "Spearman's coeff. (E: repulsion)",
+            "Spearman's coeff. (F: descending)",
             "Tortuosity",
             "Energy jump [eV]",
             "Force flips",
-            "Spearman's coeff. (Energy - attraction)",
-            "Spearman's coeff. (Force - ascending)",
+            "Spearman's coeff. (E: attraction)",
+            "Spearman's coeff. (F: ascending)",
+            "PBE energy MAE [eV]",
+            "PBE force MAE [eV/Å]",
         ]
     )
 )
@@ -146,8 +166,8 @@ def render():
             \\text{Conservation deviation} = \\left\\langle\\left| \\mathbf{F}(\\mathbf{r})\\cdot\\frac{\\mathbf{r}}{\\|\\mathbf{r}\\|} +  \\nabla_rE\\right|\\right\\rangle_{r = \\|\\mathbf{r}\\|}
             $$
 
-            - **Spearman's coeff. (Energy - repulsion)**: Spearman's correlation coefficient of energy prediction within equilibrium distance $r \\in (r_{min}, r_o = \\argmin_{r} E(r))$.
-            - **Spearman's coeff. (Force - descending)**: Spearman's correlation coefficient of force prediction within equilibrium distance $r \\in (r_{min}, r_o = \\argmin_{r} E(r))$.
+            - **Spearman's coeff. (E: repulsion)**: Spearman's correlation coefficient of energy prediction within equilibrium distance $r \\in (r_{min}, r_o = \\argmin_{r} E(r))$.
+            - **Spearman's coeff. (F: descending)**: Spearman's correlation coefficient of force prediction within equilibrium distance $r \\in (r_{min}, r_o = \\argmin_{r} E(r))$.
             - **Tortuosity**: The ratio between total variation in energy and sum of absolute energy differences between $r_{min}$, $r_o$, and $r_{max}$.
             - **Energy jump**: The sum of energy discontinuity.
             - **Force flips**: The number of sign changes.

serve/ranks/thermal-conductivity.py

@@ -11,33 +11,33 @@ table = pd.read_csv(DATA_DIR / "wte.csv")
 table.rename(
     columns={
         "method": "Model",
-        "srme": "SRME",
+        "srme": "SRME[𝜅]",
     },
     inplace=True,
 )
 
 table.set_index("Model", inplace=True)
 
-table.sort_values(["SRME"], ascending=True, inplace=True)
+table.sort_values(["SRME[𝜅]"], ascending=True, inplace=True)
 
-table["Rank"] = table["SRME"].rank(method='min').astype(int)
+table["Rank"] = table["SRME[𝜅]"].rank(method='min').astype(int)
 
 table = table.reindex(
     columns=[
         "Rank",
-        "SRME",
+        "SRME[𝜅]",
     ]
 )
 
 s = (
     table.style.background_gradient(
-        cmap="Reds", subset=["SRME"]
+        cmap="Reds", subset=["SRME[𝜅]"]
     )
     .background_gradient(
         cmap="Blues",
         subset=["Rank"],
     )
-    .format("{:.3f}", subset=["SRME"])
+    .format("{:.3f}", subset=["SRME[𝜅]"])
 )
 
 

serve/tasks/homonuclear-diatomics.py

@@ -30,9 +30,9 @@ valid_models = [
 mlip_methods = container.multiselect(
     "MLIPs",
     valid_models,
-    ["EquiformerV2(OC22)", "CHGNet", "M3GNet", "SevenNet", "MACE-MP(M)", "ORB", "eqV2(OMat)"],
+    ["MACE-MP(M)", "CHGNet", "M3GNet", "SevenNet", "ORB", "ORBv2", "eqV2(OMat)"],
 )
-dft_methods = container.multiselect("DFT Methods", ["GPAW"], [])
+dft_methods = container.multiselect("DFT Methods", ["PBE"], ["PBE"])
 
 st.markdown("### Settings")
 vis = st.container(border=True)
@@ -71,8 +71,8 @@ def get_data(mlip_methods, dft_methods):
     ]
     dfs.extend(
         [
-            pd.read_json(DATA_DIR / method.lower() / "homonuclear-diatomics.json")
-            for method in dft_methods
+            pd.read_json(DATA_DIR / "vasp" / "homonuclear-diatomics.json")
+            # for method in dft_methods
         ]
     )
     df = pd.concat(dfs, ignore_index=True)
@@ -118,63 +118,110 @@ def get_plots(df, energy_plot: bool, force_plot: bool, method_color_mapping: dic
 
             rs = rs[ind]
             es = es[ind]
-            if "GPAW" not in method:
-                es = es - es[-1]
-            else:
-                pass
+            fs = fs[ind]
+            
+            # if method not in ["PBE"]:
+            es = es - es[-1]
 
-            if "GPAW" not in method:
-                fs = fs[ind]
 
-            if "GPAW" in method:
-                xs = np.linspace(rs.min() * 0.99, rs.max() * 1.01, int(5e2))
-            else:
-                xs = rs
+            # if method in ["PBE"]:
+            #     xs = np.linspace(rs.min() * 0.99, rs.max() * 1.01, int(5e2))
+            # else:
+            xs = rs
 
             if energy_plot:
-                if "GPAW" in method:
-                    cs = CubicSpline(rs, es)
-                    ys = cs(xs)
-                else:
-                    ys = es
+                # if "GPAW" in method:
+                #     cs = CubicSpline(rs, es)
+                #     ys = cs(xs)
+                # else:
+                ys = es
 
                 elo = min(elo, max(ys.min() * 1.2, -15), -1)
-                # elo = min(elo, ys.min()*1.2, -1)
-
-                fig.add_trace(
-                    go.Scatter(
-                        x=xs,
-                        y=ys,
-                        mode="lines",
-                        line=dict(
-                            color=method_color_mapping[method],
-                            width=3,
+
+                if method in ["PBE"]:
+                    fig.add_trace(
+                        go.Scatter(
+                            x=xs,
+                            y=ys,
+                            mode="markers",
+                            line=dict(
+                                color=method_color_mapping[method],
+                                width=3,
+                            ),
+                            name=method,
                         ),
-                        name=method,
-                    ),
-                    secondary_y=False,
-                )
+                        secondary_y=False,
+                    )
+                    # xs = np.linspace(rs.min() * 0.99, rs.max() * 1.01, int(5e2))
+                    # cs = CubicSpline(rs, es)
+                    # ys = cs(xs)
+                    # fig.add_trace(
+                    #     go.Scatter(
+                    #         x=xs,
+                    #         y=ys,
+                    #         mode="lines",
+                    #         line=dict(
+                    #             color=method_color_mapping[method],
+                    #             width=3,
+                    #         ),
+                    #         name=method,
+                    #         showlegend=False,
+                    #     ),
+                    #     secondary_y=False,
+                    # )
+                else:
+                    fig.add_trace(
+                        go.Scatter(
+                            x=xs,
+                            y=ys,
+                            mode="lines",
+                            line=dict(
+                                color=method_color_mapping[method],
+                                width=3,
+                            ),
+                            name=method,
+                        ),
+                        secondary_y=False,
+                    )
 
-            if force_plot and "GPAW" not in method:
+            # if force_plot and method not in ["PBE"]:
+            if force_plot:
                 ys = fs
 
                 flo = min(flo, max(ys.min() * 1.2, -50))
 
-                fig.add_trace(
-                    go.Scatter(
-                        x=xs,
-                        y=ys,
-                        mode="lines",
-                        line=dict(
-                            color=method_color_mapping[method],
-                            width=2,
-                            dash="dashdot",
+                if method in ["PBE"]:
+                    fig.add_trace(
+                        go.Scatter(
+                            x=xs,
+                            y=ys,
+                            mode="lines+markers",
+                            line=dict(
+                                color=method_color_mapping[method],
+                                width=2,
+                                dash="dashdot",
+                            ),
+                            name=method,
+                            showlegend=not energy_plot,
                         ),
-                        name=method,
-                        showlegend=not energy_plot,
-                    ),
-                    secondary_y=True,
-                )
+                        secondary_y=True,
+                    )
+                else:
+                    fig.add_trace(
+                        go.Scatter(
+                            x=xs,
+                            y=ys,
+                            mode="lines",
+                            line=dict(
+                                color=method_color_mapping[method],
+                                width=2,
+                                dash="dashdot",
+                            ),
+                            name=method,
+                            showlegend=not energy_plot,
+                        ),
+                        secondary_y=True,
+                    )
 
         name = f"{symbol}-{symbol}"
 
@@ -187,6 +234,7 @@ def get_plots(df, energy_plot: bool, force_plot: bool, method_color_mapping: dic
                 y=1,
                 yanchor="top",
                 bgcolor="rgba(0, 0, 0, 0)",
+                # traceorder='reversed',
                 # entrywidth=0.3,
                 # entrywidthmode='fraction',
             ),

serve/tasks/thermal-conductivity.py

@@ -26,20 +26,20 @@ table = pd.read_csv(DATA_DIR / "wte.csv")
 table.rename(
     columns={
         "method": "Model",
-        "srme": "SRME",
+        "srme": "SRME[𝜅]",
     },
     inplace=True,
 )
 
 table.set_index("Model", inplace=True)
 
-table.sort_values(["SRME"], ascending=True, inplace=True)
+table.sort_values(["SRME[𝜅]"], ascending=True, inplace=True)
 
 s = (
     table.style.background_gradient(
-        cmap="Reds", subset=["SRME"]
+        cmap="Reds", subset=["SRME[𝜅]"]
     )
-    .format("{:.3f}", subset=["SRME"])
+    .format("{:.3f}", subset=["SRME[𝜅]"])
 )
 
 st.dataframe(

tests/test_external_calculators.py

@@ -12,6 +12,9 @@ def test_calculate(model: MLIPEnum):
     if model.name == "ALIGNN":
         pytest.xfail("ALIGNN has poor file download mechanism")
 
+    if model.name == "ORB":
+        pytest.xfail("Orbital Materials deprecated the model a month after its premature release in favor of ORBv2")
+
     try:
         calc = MLIPEnum[model.name].value()