From b2ec64c7ce5bbb0f34731ee35f3376e8dbb4b7cc Mon Sep 17 00:00:00 2001
From: Simonas <20096648+simjak@users.noreply.github.com>
Date: Mon, 18 Dec 2023 10:26:02 +0200
Subject: [PATCH] wip
---
docs/examples/function_calling.ipynb | 124 +++++++++++++++++++++++++++
docs/examples/test.ann | Bin 0 -> 64764 bytes
poetry.lock | 15 +++-
pyproject.toml | 1 +
4 files changed, 138 insertions(+), 2 deletions(-)
create mode 100644 docs/examples/test.ann
diff --git a/docs/examples/function_calling.ipynb b/docs/examples/function_calling.ipynb
index 98beb735..29ca749b 100644
--- a/docs/examples/function_calling.ipynb
+++ b/docs/examples/function_calling.ipynb
@@ -516,6 +516,130 @@
"call(query=\"What is the tech news in the Lithuania?\", functions=tools, router=router)\n",
"call(query=\"Hi!\", functions=tools, router=router)"
]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 1,
+ "metadata": {},
+ "outputs": [
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "[0.23461914, 0.50146484, -0.048828125, 0.13989258, -0.18029785]\n"
+ ]
+ }
+ ],
+ "source": [
+ "import cohere\n",
+ "import os\n",
+ "\n",
+ "co = cohere.Client(api_key=os.environ[\"COHERE_API_KEY\"])\n",
+ "texts = [\n",
+ " \"Hello from Cohere!\",\n",
+ " \"مرØبًا من كوهير!\",\n",
+ " \"Hallo von Cohere!\",\n",
+ " \"Bonjour de Cohere!\",\n",
+ " \"¡Hola desde Cohere!\",\n",
+ " \"Olá do Cohere!\",\n",
+ " \"Ciao da Cohere!\",\n",
+ " \"您好，æ¥è‡ª Cohereï¼\",\n",
+ " \"कोहेरे से नमसà¥à¤¤à¥‡!\",\n",
+ "]\n",
+ "response = co.embed(texts=texts, model=\"embed-multilingual-v2.0\")\n",
+ "embeddings = response.embeddings # All text embeddings\n",
+ "print(embeddings[0][:5]) # Print embeddings for the first text"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 2,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "response = co.detect_language(\n",
+ " texts=[\n",
+ " \"\"\"Lenkijos įstatymų leidÄ—jai pavedÄ— buvusiam Europos Vadovų Tarybos vadovui Donaldui Tuskui suformuoti VyriausybÄ™ po aÅ¡tuonerius metus trukusio deÅ¡iniųjų valdymo. Ar tai reiÅ¡kia, kad baigsis Lenkijos pykÄiai su Ukraina ir Europos SÄ…junga? Ar tai gera žinia ir Lietuvai? Kokių pokyÄių apskritai reikÄ—tų tikÄ—tis? Apie tai buvo diskutuojama Žinių radijo laidoje „Aktualusis interviu\"\"\"\n",
+ " ]\n",
+ ")\n",
+ "\n",
+ "response\n",
+ "texts = [\n",
+ " \"\"\"Lenkijos įstatymų leidÄ—jai pavedÄ— buvusiam Europos Vadovų Tarybos vadovui Donaldui Tuskui suformuoti VyriausybÄ™ po aÅ¡tuonerius metus trukusio deÅ¡iniųjų valdymo. Ar tai reiÅ¡kia, kad baigsis Lenkijos pykÄiai su Ukraina ir Europos SÄ…junga? Ar tai gera žinia ir Lietuvai? Kokių pokyÄių apskritai reikÄ—tų tikÄ—tis? Apie tai buvo diskutuojama Žinių radijo laidoje „Aktualusis interviu\"\"\"\n",
+ "]"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 3,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "embeds = co.embed(\n",
+ " texts=texts, model=\"embed-multilingual-v2.0\", input_type=\"search_document\"\n",
+ ").embeddings"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 4,
+ "metadata": {},
+ "outputs": [
+ {
+ "data": {
+ "text/plain": [
+ "True"
+ ]
+ },
+ "execution_count": 4,
+ "metadata": {},
+ "output_type": "execute_result"
+ }
+ ],
+ "source": [
+ "from annoy import AnnoyIndex\n",
+ "import numpy as np\n",
+ "\n",
+ "# Create the search index, pass the size of embedding\n",
+ "search_index = AnnoyIndex(np.array(embeds).shape[1], \"angular\")\n",
+ "\n",
+ "# Add all the vectors to the search index\n",
+ "for i in range(len(embeds)):\n",
+ " search_index.add_item(i, embeds[i])\n",
+ "search_index.build(10) # 10 trees\n",
+ "search_index.save(\"test.ann\")"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "query = \"What is the tallest mountain in the world?\"\n",
+ "input_type_query = \"search_query\"\n",
+ "\n",
+ "# Get the query's embedding\n",
+ "query_embed = co.embed(\n",
+ " texts=[query], model=\"embed-multilingual-v2.0\", input_type=input_type_query\n",
+ ").embeddings\n",
+ "\n",
+ "# Retrieve the nearest neighbors\n",
+ "similar_item_ids = search_index.get_nns_by_vector(\n",
+ " query_embed[0], 10, include_distances=True\n",
+ ")\n",
+ "# Format the results\n",
+ "query_results = pd.DataFrame(\n",
+ " data={\n",
+ " \"texts\": df.iloc[similar_item_ids[0]][\"text\"],\n",
+ " \"distance\": similar_item_ids[1],\n",
+ " }\n",
+ ")\n",
+ "\n",
+ "\n",
+ "print(f\"Query:'{query}'\\nNearest neighbors:\")\n",
+ "print(query_results) # NOTE: Your results might look slightly different to ours"
+ ]
}
],
"metadata": {
diff --git a/docs/examples/test.ann b/docs/examples/test.ann
new file mode 100644
index 0000000000000000000000000000000000000000..85db20d54995bcd7f1199d1816a069e14558aa40
GIT binary patch
literal 64764
zcmeH~|C3d98ONUmgaO-F1q8R$gRp{#@<lLf_uhMq-Oz9gtyF^Gkj)q{go87rsLc@;
zVNnbSaR8w!B1tT)F@V6`d+x`AbeLq+utLl}+BgV=l7{$YkX{G>jPw5C>^bLip3n2V
zpV#wzK6gJ=X7<vw?~dVS7R@W0?@}>4oVVceycMT%^-HX{D{t|M#N03P7QVstW@7dV
z?^ow6Z037DvEW3`ifzopHr&j8Mc!-y-=5&!mY9Eu=N9hA6N|c5%zqD?5`Hf%SU8e7
zmnIfIoHO5AF!wJ0?<iYwDSjWsW-^+tELhM^ya~L|#K-N}?q%MwoViPxvp;Xevpla#
zEIu!9elGWp1`D3#d2PjlfAKvmo12@n;7H!0i8+gF(Y%K5w^b}UzhM3uG+k1#h}^>a
zVexF<TpxT)!q>}u>x713*e7L+(eJL{d1-^WNBQ<5`Mu728$RI^G!pL%-h1Nr9=@%?
zXMMrqK?RtS&nD)ymo0dY@zL0hz#n<|FUb4n#Cwc6yUFkA#QbEw^DR1^hgZeoBHz~#
z_iWj0J~_|g`_8h3+lUME_+ZKWR`|d)wRh8)w+YQBsl`AT{t@2;iPu!Iu$tVvp=)_!
z?hs?|qq#rxsEw_{|5Ws@$M0X!LhXZXFdV}C``|MHyLspz&NT|{f54^{A9Xpi7UnO(
z=R)ke!rJk?9)IJ>V<|Shh}nbZ)g=qAq#iSfJEvf_`pbF#nZ)dQa$kY(E%^PA9RA1s
zKfL#?m>-O;`S^Sb|LM1j&{j(?Jw<PzBRa(uaPjV{sDD@d9>r!kW4AKajXUEpHIJU7
z4gq?|KiHdCG!$L<^5k8dji%lm7B9qZ42&0}=}E4~uuXgABcA^Pe|p2UbXd?G-S0G5
zxT?XdF0tS`G)*S&#n@kk4eQB2!1F(d^C{O2Ws9bxufAfDW4t#uPt!y7Fskn`cRsnH
z%hP}ENBCwfMN_%c4{;6g)A2`$^8vNkhTQ<_&YB3OV}r)%AmeMXnbTm!L(H9z)>AO+
z3rG4Yo=MDWiT!V$>A|oM{qig9$1}c|7@NqQH5{VH&p`WfG*#0>X!X;m>u21b;`>DW
z4&dGluZt`62R7)AtNGp&pKvSQ&UfOu5or4<T)OjK#rW?St0jji=-?jFXYM$8;3s?>
z+utM>?jxr+ut~Y^K|?#^Z{;k4NAZ{F>Ib)3^;V=O{T}L>*0sdE;pm}0;b43ZM&qma
zX1#=uP~VaG`ZfIjnKyu&KfoG!8#X&(cADHi$H$bCh4-Q9TWI+@>z}z1IR~TBPcH`3
z__ly}$MOC5#2h^xl8<K{7C&HY8S5m6&I=gpMBm2rMjd|cVqOb%-ozYwzPJNj>{o7P
z#SG1U81qBy(B@k4u?M@g=uF4bz3vk9+zsEq!G+ot8_D%+#5l~_>&->(whyh{sKZvi
z^(OYsj7_3vzJkUUwBH7YBiP`_zlxt!O9Q#3dqF?mSyOQd?d-#TH}`Z+_n@{zV8cG|
z*wdnQ*skRJAoTtWcAfEk47>m0w<)yf06r!V^F?BH=1T1kqPIU<UPHrJe4!;8!}AM_
z4Iy{>DwvC=8m>N!oyS<(SDrYRQRg=1A8jx{9j){5`4Hnp;__SK*#{%|TTRZ~8T>3l
z*DJ)l8V#(&Xg~bF&%D=DPV~e|*!E@5y_Nms8zr-o)ILe|@E#GL-jAre;WK(WvG5>z
z(Ca@zLlrjdj8{{WA21#eYcsy+0s9o2%h_MBv6Cf>+3!3)T^e%(npvyCv(&$WwtZZ1
z^Kc1ROYWMo1y^8ejM3}y4QTx>vCH%adstjYoJOv*_*qS^^tUbH>W)8Z5NxOB_mgi=
z=G^TpxSrZQ1fO1MPxJgZnlD2iePR1L%>NiabI4~GKI@r-X4i<#dbEtHn13IZ<MH<{
zbJBfrI~sbSwT?NPnbU;bEat%^+)=ie{oLP6PHFDMihqwL59jajz6#s^=&0s9zxV0d
zjdwGTHF55@XA-%NW6q5gbI0~r(3$x3M%2jM75H67%%w2C5$4n(fPK7%oY<FaBlE7S
zSnxya$R~OyXMP=A-h&5rZXt77*KP}W^LynUCWld6iz*g|=xXKrC*;8%6ApwIYcSqk
zHXDmA`(kttu`UkH4zd5#gyt*Q&n7PWdTJl7W3Nc-@@K9tu)33ei_!i$b{px})9BrT
zu94(45F2{mnpiVKiASs;UEd?%bcVPuV@F-XyI}P^d8Hiq?e=FFr`GWa=J|Syk7Bzd
zv~Vx}x{-G+>l5zL1?Wv}Tj6{lv1mBj$D}OCm)|5e08aO!t0%r*<Jy1?Yc0M7KU@L-
zw^*LB(d5y=9Q1g4GNL~M;yU)fXlK3I1~|~$&cSH`{vM>(w~*Ijc-}$p9w+7yn3l-v
zP4tXoKKoU8Eitc3H4~Hmi|N^7GrjWweurVVgPP4I=dH|7d!-4z&9JD&?{nB+1oIE6
zCG*ee8iegP3+7oz#iqn8{a=mW4Sy3p$=&**?HY8_fAI|dryGmUN%)?C?(X<xJ=$Ju
z#$l7PxE>AkxutvcS7G@v`rsQhpsx)UJsQ}1@x^<*pLs5|&#``)KNb5U#O=a;85*WA
z)=s>Y*e~M!V`8$8y7d25?1Ck5*vPdIZZFWwdF&3u2~8pEC+vi;&$#~tJrn8u{^XeU
zF20HvV*gQc?th}FnZ4~!IPw2O%svu-8y|n-nZ3hy(J!a)QIlAFGjmxNQ8WICpYHF$
z&|KBILxBE>fCz|y2#A0Ph=2%)fCyxbfX)J0tGMzN0TB=Z5fA|p5CIVofiDT@ETB#h
z0TB=Z5fA|p5CIVofou}cSs<HrSJomR0wN#+A|L`HAOa$wvw(U)1VlgtL_h>YKm<fU
z1hPp$XMt?iU0I8O2#A0Ph=2%)fCz|y&I0NI5fA|p5CIVo0TB=Z5y&P1odvR4cV#UC
zA|L`HAOa#F0wN#+It!==L_h>YKm<fU1VlgtL?D|4bQZ{F-IcWnh=2%)fCz|y2#A0P
z=q#Wf5CIVo0TB=Z5fA|p5P@tG&{-gxbywCRAOa#F0wN#+A|L`HptFE_Km<fU1Vlgt
zL_h>YKm@W$Kxct$)?HbPfCz|y2#A0Ph=2%)fX)Ky0TB=Z5fA|p5CIVo0TIY10i6Z1
zS$AbE0wN#+A|L`HAOa#F0y+z*2Sh*wL_h>YKm<fU1VkX41aub2X5E#w2#A0Ph=2%)
yfCz|y2<R-J9uNT$5CIVo0TB=Z5fFiF63|&7n{`*#A|L`HAOa#F0wN#+BJh6`iW`0a
literal 0
HcmV?d00001
diff --git a/poetry.lock b/poetry.lock
index f5d58647..6cd649bb 100644
--- a/poetry.lock
+++ b/poetry.lock
@@ -1,4 +1,4 @@
-# This file is automatically @generated by Poetry 1.6.1 and should not be changed by hand.
+# This file is automatically @generated by Poetry 1.5.1 and should not be changed by hand.
[[package]]
name = "aiohttp"
@@ -110,6 +110,17 @@ files = [
[package.dependencies]
frozenlist = ">=1.1.0"
+[[package]]
+name = "annoy"
+version = "1.17.3"
+description = "Approximate Nearest Neighbors in C++/Python optimized for memory usage and loading/saving to disk."
+optional = false
+python-versions = "*"
+files = [
+ {file = "annoy-1.17.3-cp310-cp310-macosx_11_0_arm64.whl", hash = "sha256:c33a5d4d344c136c84976bfb2825760142a8bb25335165e24e11c9afbfa8c2e9"},
+ {file = "annoy-1.17.3.tar.gz", hash = "sha256:9cbfebefe0a5f843eba29c6be4c84d601f4f41ad4ded0486f1b88c3b07739c15"},
+]
+
[[package]]
name = "anyio"
version = "4.1.0"
@@ -2203,4 +2214,4 @@ testing = ["big-O", "jaraco.functools", "jaraco.itertools", "more-itertools", "p
[metadata]
lock-version = "2.0"
python-versions = "^3.10"
-content-hash = "3300c77d6b6fab3faca403e2f3064a23e9f5ddcd34d63cad42d39b94b1ae5c2b"
+content-hash = "8d9e62152a77382c673b01ba8279001795093ae90fc9d67e4073f4c1b5b5c10e"
diff --git a/pyproject.toml b/pyproject.toml
index b041516c..dc0b853e 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -30,6 +30,7 @@ pytest-mock = "^3.12.0"
pytest-cov = "^4.1.0"
pytest-xdist = "^3.5.0"
mypy = "^1.7.1"
+annoy = "^1.17.3"
[build-system]
requires = ["poetry-core"]
--
GitLab